With the development of deep learning, several Graph Neural Networks (GNN)-based approaches have been utilized for text classification. However, GNNs encounter challenges in capturing contextual text information within a document sequence. To address this, a novel text classification model RB-GAT is proposed by combining RoBERTa-BiGRU embedding and a multi-head Graph ATtention Network (GAT). First, the pre-trained RoBERTa model is exploited to learn word and text embeddings in different contexts. Second, the Bidirectional Gated Recurrent Unit (BiGRU) is employed to capture long-term dependencies and bidirectional sentence information from the text context. Next, the multi-head graph attention network is applied to analyze this information, which serves as a node feature for the document. Finally, the classification results are generated through a Softmax layer. Experimental results on three benchmark datasets demonstrate that our method can achieve an accuracy of 71.48%, 98.45%, and 80.32% on Ohsumed, R8, and MR, which is superior to the existing nine text classification approaches.

Sentiment analysis plays a crucial role in understanding customer feedback, guiding product development, and informing business decisions. This paper introduces term sentiment entropy (TSE), a novel weighting method that leverages the distribution of sentiment labels associated with words to enhance text classification accuracy. TSE complements traditional TF-IDF techniques by capturing the nuances of sentiment variation across different contexts. Experiments across diverse public datasets demonstrate TSE's potential to improve sentiment analysis performance, especially in capturing subtle sentiment shifts and adapting to specific domains. While computational cost and label quality pose challenges, TSE offers a promising avenue for refining sentiment analysis and opening new research frontiers.

We describe a multi-resolution approach for audio classification and illustrate its application to the open data set for environmental sound classification. The proposed approach utilizes a multi-resolution based ensemble consisting of targeted feature extraction of approximation (coarse scale) and detail (fine scale) portions of the signal under the action of multiple transforms. This is paired with an automatic machine learning engine for algorithm and parameter selection and the LSTM algorithm, capable of mapping several sequences of features to a predicted class membership probability distribution. Initial results show an improvement in multi-class classification accuracy.

Open-vocabulary learning has recently gained prominence as a means to enable image segmentation for arbitrary categories based on textual descriptions. This advancement has extended the applicability of segmentation systems to a broader range of generally purpose scenarios. However, current methods often revolve around specialized architectures and parameters tailored to specific segmentation tasks, resulting in a fragmented landscape of segmentation models. In response to these challenges, we introduce OVAMTSeg, a versatile framework designed for Open-Vocabulary and Multitask Image Segmentation. OVAMTSeg harnesses adaptive prompt learning to empower the model to capture category-sensitive concepts, enhancing its robustness across diverse multi-task and scenario contexts. Text prompts are employed to effectively capture semantic and contextual features of the text, while cross-attention and cross-modal interactions enable the fusion of image and text features. Furthermore, a transformer-based decoder is incorporated for dense prediction. Extensive experimental results underscore the effectiveness of OVAMTSeg, showcasing its state-of-the-art performance and superior generalization capabilities across three segmentation tasks. Notable achievements include a 47.5 mIoU in referring expression segmentation, 51.6 mIoU on Pascal-VOC with four unseen classes, 46.6 mIoU on Pascal-Context in zero-shot segmentation, 65.9 mIoU on Pascal-5i, and 35.7 mIoU on COCO-20i datasets for one-shot segmentation.

In this paper, we present a novel approach for text independent phone-to-audio alignment based on phoneme recognition, representation learning and knowledge transfer. Our method leverages a self-supervised model (wav2vec2) fine-tuned for phoneme recognition using a Connectionist Temporal Classification (CTC) loss, a dimension reduction model and a frame-level phoneme classifier trained thanks to forced-alignment labels (using Montreal Forced Aligner) to produce multi-lingual phonetic representations, thus requiring minimal additional training. We evaluate our model using synthetic native data from the TIMIT dataset and the SCRIBE dataset for American and British English, respectively. Our proposed model outperforms the state-of-the-art (charsiu) in statistical metrics and has applications in language learning and speech processing systems. We leave experiments on other languages for future work but the design of the system makes it easily adaptable to other languages.

Successful geophysical exploration works in Gucheng-Yaxi area located in Gaochun District, Jiangsu province, China, have been limited partly due to complex geological condition of the area and high artificial noise in data acquired using electrical and electromagnetic methods. In this work, we deployed the new anti-interference spread spectrum induced polarization method (SSIP) and audio magnetotellurics (AMT) method to detect a copper-polymetallic deposit in the area. Two-dimensional (2D) inversion results at Gucheng-Yaxi Area revealed a high chargeability anomalous zone on SSIP profile that coincided with a zone of moderate resistivity located between two resistor bodies on AMT profile. A follow up 1200 meters drill hole was established at this high chargeability-moderate resistivity zone which encountered polymetallic (copper, lead, zinc, gold and silver) mineralization at the depth ≥ 400 m. Drill hole data analysis showed that mineralization occurs interspaced in marble rock mass at varies depth. Furthermore, several low resistivity-weak chargeability sections were revealed and inferred attributed to Cretaceous sediments and faults. These faults are thought to have played critical role in the polymetallic mineralization genesis. In summary, this work demonstrated the successful of application of SSIP and AMT at detecting a metallic deposit in an area with high artificial noise. Hence the geophysical prospection potential of Gucheng-Yaxi area is great.

Keywords: Linear differential arrays; beam steering; frequency-invariant beamforming.

Embedding representation models characterize each word as a vector of numbers with a fixed length. These models have been used in tasks involving text classification, such as recommen- dation and question-answer systems. Semantic relationships are words with a relationship between them providing a complete idea to a text. Therefore, it is hypothesized that an embedding model involving semantic relationships will provide better performance for tasks that use them. This paper presents three embedding models based on semantic relations extracted fromWikipedia to classify texts. The synonym, hyponym, and hyperonym semantic relationships were the ones considered in this work since previous experiments have shown that they are the ones that provide the most semantic knowledge. Lexical-syntactic patterns present in the literature were implemented and subsequently applied to the Wikipedia corpus to obtain the semantic relationships present in it. Several semantic relationships are used in different models: synonymy, hyponym-hyperonym, and a combination of the first two. A convolutional neural network was trained for text classification to evaluate the performance of each model. The results obtained were evaluated with the metrics of precision, accuracy, recall, and F1-measure. The best values obtained with the second model were accuracy of 0.79 for the 20-Newsgroup corpus. F1-measure and recall of 0.87 respectively for the Reuters corpus.

This research study focuses on addressing practical optimization challenges in audio-visual content using advanced computational intelligence methods, with a specific emphasis on privacy-preserving issues. The research will be presented as a proposal and can be conducted based on the provided ideas. Computational intelligence encompasses various approaches, including evolutionary algorithms, learning methods, fuzzy systems, and neural networks, which have become essential tools for tackling complex engineering optimization problems due to computational expenses, dimensional complexities, and interdisciplinary integration. The research objectives include a thorough investigation of optimization challenges in audio-visual content, the delineation of problem characteristics, and the development of relevant methodologies based on these specifications. The main goals of the study, titled "Advanced Computational Intelligence Assisted Less-Expensive Privacy-Preserving Optimization," encompass the development of a metamodel-based learning algorithm integrated with evolutionary computation for privacy-preserving optimization, less-expensive signal watermarking optimization using hybrid metamodeling and evolutionary algorithms, and a closed-loop robust intelligent PID control system applied in real-time dynamic signal watermarking. Additional objectives may be defined as the research progresses and further background review is conducted.

Models of subsurface structures are important for successful deposit exploration but are challenged by the need to integrate data from different geophysical methods. In the present study, we evaluated a method of joint inversion in which audio-magnetotelluric (AMT) and dual frequency induced polarization (DFIP) data sets are inverted simultaneously to produce a consistent 2D resistivity model to show a clear image of subsurface structures. To achieve the objectives, we conducted AMT and DFIP surveys first on the same survey line within the Dongjun lead-zinc deposit in inner Mongolia by developing 31 AMT survey sites with an interstation separation of 40 m on a 1440 m survey track and operated in fifty-three frequencies in the range 1–10,400 Hz to record the resistivity distribution of subsurface to depths exceeding 800 m. The same survey setup up was applied to the DFIP method by utilizing the SQ-3C model with a pole-dipole array configuration and conducted measurements at frequencies of 4 Hz and 4/13 Hz. We first separately inverted AMT and DFIP measurements. The two-dimensional (2D) model obtained from AMT data revealed distinct low resistivity anomalies in the middle of the 2D inversion model. In contrast, the DFIP inversion model showed a high resistive body in the same region. In response to the discrepancies observed in the separate 2D inversion models, we implemented a joint inversion for both the AMT and DFIP datasets. The joint inversion resistivity model shows surficial conducting bodies and a high conductive body along the profile with relatively high Percent Frequency Effect (PEF) indicating high chargeability. The final joint inversion resistivity model clearly images the large silica alteration zone and the Pb-Zn mineralization. This study demonstrates the feasibility of a joint inversion methodology and highlights the value of integrating geophysical methods through joint inversion for enhanced characterization and exploration of lead-zinc ores.

The emergence of Artificial Intelligence (AI) driven Keyword Spotting (KWS) technologies has revolutionized human to machine interaction. Yet, the challenge of end-to-end energy efficiency, memory footprint and system complexity of current Neural Network (NN) powered AI-KWS pipelines has remained ever present. This paper evaluates KWS utilizing a learning automata powered machine learning algorithm called the Tsetlin Machine (TM). Through significant reduction in parameter requirements and choosing logic over arithmetic based processing, the TM offers new opportunities for low-power KWS while maintaining high learning efficacy. In this paper we explore a TM based keyword spotting (KWS) pipeline to demonstrate low complexity with faster rate of convergence compared to NNs. Further, we investigate the scalability with increasing keywords and explore the potential for enabling low-power on-chip KWS.

Multimodal remote sensing data classification can enhance the model’s ability to distinguish land features through multimodal data fusion. In this context, how to help models understand the relationship between multimodal data and target tasks is the focus of researchers. Inspired by human feedback learning mechanisms, causal reasoning mechanisms, and knowledge induction mechanisms, this paper integrates causal learning, reinforcement learning, and meta learning into a unified remote sensing data classification framework and proposed the causal meta-reinforcement learning (CMRL). First, based on feedback learning mechanisms, we have overcome the limitations of traditional implicit optimization of fusion features and customized a reinforcement learning environment for multimodal remote sensing data classification tasks. Through feedback interactive learning between agents and the environment, we help them understand the complex relationships between multimodal data and labels, thereby achieving full mining of multimodal complementary information. Second, based on the causal inference mechanism we designed causal distribution prediction action, classification reward, and causal intervention reward, capturing pure causal factors in multimodal data and cutting off false statistical associations between non-causal factors and class labels. Finally, based on the knowledge induction mechanism, we designed a bi-layer optimization mechanism based on meta-learning. By constructing a meta training task and meta validation task simulation model in the generalization scenario of unseen data, we helped the model induce cross-task shared knowledge, thereby improving its generalization ability for unseen multimodal data. The experimental results on multiple sets of multimodal datasets show that the proposed method achieved state-of-the-art performance in multimodal remote sensing data classification tasks.

Topic discovery is finding the main idea of large amounts of textual data. It indicates the recurring topics in the documents, allowing an overview of the texts. Current topic discovery models receive the texts, with or without pre-processing of Natural Language Processing. The processing consists of stopwords removal, text cleaning and normalization (lowercase conversion). A topic discovery model that receives texts with or without processing generates general topics since the input data is many uncategorized texts. The general topics do not offer a detailed overview of the input texts, and manual text categorization is a time-consuming and tedious task. Accordingly, it is necessary to integrate an automatic text classification task in the topic discovery process to obtain specific topics with their top words that contain relevant relationships based on belonging to a class. Text classification performs a word analysis that makes up a document to decide what class or category is being identified; then, integrating the text classification before a topic discovery process will provide latent topics depicted by top words with a high coherence in each topic based on the previously obtained classes. Therefore, this paper exposes a approach that integrates text classification into topic discovery from large amounts of English textual data, such as 20-Newsgroup and Reuters corpora. The text classification is accomplished with a Convolutional Neural Network(CNN) incorporating three embedding models based on semantic relationships. The topic discovery over categorized texts is realized with Latent Dirichlet Analysis(LDA), Probabilistic Latent Semantic Analysis(PLSA), and Latent Semantic Analysis(LSA) algorithms. An evaluation process was performed based on the normalized topic coherence metric. The 20-Newsgroup corpus was classified, and twenty topics with ten top words were discovered for each class, obtaining 0.1723 normalized topic coherence when applying LDA, 0.1622 with LSA, and 0.1716 with PLSA. The Reuters corpus was also classified, obtaining 0.1441 normalized topic coherence when applying the LDA algorithm to obtain 20 topics for each class.

Against the background of the sky, imaging and perception of aircraft are cru- cial. Among these, the location of aircrafts is not a very difficult problem in general, because that deep learning algorithms and object detection models based on convolutional neural network(CNN) are ever-evolving. In order to have a better identification of various aircrafts, or to accurately locate different types of flying targets in the sky, it becomes significant to distinguish and recognize different types of aircrafts. One question is that, distinguishing and recognizing between sub-categories of aircrafts pose great challenges. Although fine-grained recognition focuses on exploring and studying such problems, aircrafts under different sub-categories and granularities lead us to rethink the application of features. We noticed that features in swin-transformer demonstrates the under- standing and processing of images, fully showcasing the encoding and indexing of information. Through this research and proposed approach, we discovered a better understanding of features encoding and use, which are inspired by the feature encoding and computation in swin-transformer. In our paper, our ap- proach has achieved effects on features encoded graphically, manifested from the architecture design and convolutional neural network computation, and outper- forms other famous fine-graiend classification models on this issue. Not only approach we proposed has demonstrated superior performance in fine-grained aicraft classification, but also the mechanisms of the feature encoding under different sample space partitions is revealed for this issue, which provides a re- search for the representation of aircraft fineness characteristics. The relationship between representation orientation of aircraft feature information under various grained divisions, shows that the recognition of different categories according to man-made, such as aircraft, can be achieved by means of the feature represen- tations studied in this paper, for more specific defined classification tasks and various man-made targets partition criteria, which may influences the princi- ple of design for calculation and feature extraction in fine-grained classification models.

In the realm of the five-category classification endeavor, there has been limited exploration into applied techniques for classifying Arabic text. These methods have primarily leaned on singletask learning, incorporating manually crafted features that lack robust sentence representations. Recently, the Transformer paradigm has emerged as a highly promising alternative. However, when these models are trained using single-task learning, they often face challenges in achieving outstanding performance and generating robust latent feature representations, especially when dealing with small datasets. This issue is particularly pronounced in the context of the Arabic dialect, which has a scarcity of available resources. Given these constraints, this study introduces an innovative approach to dissecting sentiment in Arabic text. This approach combines Inductive Transfer (INT) with the Transformer paradigm to augment the adaptability of the model and refine the representation of sentences. By employing self-attention SE-A and feed-forward sub-layers as a shared Transformer encoder for both the five-category and three-category Arabic text classification tasks, this proposed model adeptly discerns sentiment in Arabic dialect sentences. The empirical findings underscore the commendable performance of the proposed model, as demonstrated in assessments of the Hotel Arabic-Reviews Dataset, the Book Reviews Arabic Dataset, and the LARB dataset.

Audio classification using deep learning models, essential for applications like voice assistants and music analysis, faces challenges when deployed on edge devices due to their limited computational resources and memory. Achieving a balance between performance, efficiency, and accuracy is a significant obstacle in optimizing these models for such constrained environments. In this investigation, we evaluate diverse deep learning architectures, including Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM), for audio classification tasks on the ESC 50, UrbanSound8k, and Audio Set datasets. Our empirical findings indicate that Mel Spectrograms outperform raw audio data, attributing this enhancement to their synergistic alignment with advanced image classification algorithms and their congruence with human auditory perception. To address the constraints of model size, we apply model compression techniques, notably magnitude pruning, Taylor Pruning, and 8-bit quantization. The research demonstrates that a hybrid pruned model achieves a commendable accuracy rate of 89 percent, which, although marginally lower than the 92 percent accuracy of the uncompressed CNN, strikingly illustrates an equilibrium between efficiency and performance. Subsequently, we deploy the optimized model on Raspberry Pi 4 and NVIDIA Jetson Nano platforms for audio classification tasks. These findings highlight the significant potential of model compression strategies in enabling effective deep learning applications on resource-limited devices, with minimal compromise on accuracy.

In an era where the proliferation of large language models blurs the lines between human and machine-generated content, discerning text authenticity is paramount. This study investigates transformer-based language models—BERT, RoBERTa, and DistilBERT—in distinguishing human-written from machine-generated text. By leveraging a comprehensive corpus, including human-written text from sources such as Wikipedia, WikiHow, various news articles in different languages, and texts generated by OpenAI's GPT-2, we conduct rigorous comparative experiments. Our findings highlight the superior effectiveness of ensemble learning models over single classifiers in this critical task. This research underscores the versatility and efficacy of transformer-based methodologies for a wide range of natural language processing applications, significantly advancing text authenticity detection systems. The results demonstrate competitive performance, with the transformer-based method achieving an F-score score of 0.83 with RoBERTa-large (monolingual) and 0.70 with DistilBERT-base-uncased (multilingual).

Cloud-induced atmospheric extinction and occlusion have a major effect on the effectiveness and quality of telescope observations. Real-time cloud cover distribution and long-term statistical data are essential for astronomical siting and telescope operations. At ground-based astronomical telescope sites, cloud cover distribution is currently analyzed using manual observation methods. However, the main disadvantages of manual observation methods are human subjective, heavy workloads and poor real-time performance. Therefore, a real-time automatic cloud images classification method is desperately needed. This paper presents a new cloud identification method, which is named PSO+XGBoost model by combining eXtreme Gradient Boosting (XGBoost) and Particle Swarm Optimization (PSO). The entire cloud image is divided into 37 sub-regions in order to more precisely identify the distribution of the clouds. 19 features are then extracted, including the sky background, star density, lighting conditions, and subregion grayscale values. The experimental results have shown that the overall classification accuracy of 96.91% and our model is able to outperform several state-of-the-art baseline methods. Our approach achieves high accuracy in comparison with the manual observation methods. Moreover, this method meets telescope real-time scheduling requirements.

Motion related human activity recognition using wearable sensors can potentially enable various useful daily applications. So far, most studies view it as a stand-alone mathematical classification problem without considering the physical nature of human motions. Consequently, they suffer from data dependencies and encounter the dimension disaster problem and the over-fitting issue, and their models are never human-readable. In this study, we start from a deep analysis on natural physical properties of human motions, and then propose a useful feature selection method to quantify each feature's classification contribution capability. On one hand, the "dimension disaster" problem can be avoid to some extent, due to the affined dimension of key features; On the other hand, over-fitting issue can be depressed since the knowledge implied in human motions are nearly invariant, which compensates the possible data inadequacy. The experiment results indicate that the proposed method performs superior to those adopted in related works, such as decision tree, k-NN, SVM, neural networks.

The existing segmentation-based scene text detection methods mostly need complicated post-processing, and the post-processing operation is separated from the training process, which greatly reduces the detection performance. The previous method, DBNet successfully simplified post-processing and integrated the post-processing into a segmentation network. However, the training process of the model took a long time for 1200 epochs and the sensitivity to texts of various scales was lacking, leading to some text instances being missed. Considering the above two problems, we design the text detection Network with Binarization of Hyperbolic Tangent(HTBNet). First of all, we propose Binarization of Hyperbolic Tangent (HTB), optimized along with which, the segmentation network can expedite the initial convergent speed by reducing the amount of epochs from 1200 to 600. Because features of different channels in the same scale feature map focus on the information of different regions in the image, to better represent the important features of all objects in the image, we devise the Multi-Scale Channel Attention(MSCA). Meanwhile considering that multi-scale objects in the image cannot be simultaneously detected, we propose a novel module named Fused Module with Channel and Spatial(FMCS), which can fuse the multi-scale feature maps from channel and spatial dimension. Finally we adopt cross entropy as the loss function, which measures the difference between predicted values and ground truths. The experimental results show that HTBNet compared with lightweight models has achieved competitive performance and speed on Total-Text(F-measure:86.0%, FPS:30) and MSRA-TD500 (F-measure:87.5%, FPS:30).

Convolutional Neural Networks (CNNs) present drawbacks for modeling geometric transformations, such as scaling and rotation, caused by the convolution operation’s locality. Deformable convolution (DCON), a mechanism that substitutes standard convolution, increasing the receptive field to capture relevant features, is a promising approach to solve this drawback and improve the robustness of CNNs. However, the optimal way to replace the standard convolution with its deformable counterpart in a CNN model is unclear. In this study, we clarify this aseveration by conducting several experiments using deformable convolutions applied in the layers that conform a small four-layer CNN model. We also use deformable convolutions on the four-layers of several ResNet CNNs with depths 18, 34, 50, and 101. The models were tested in binary balanced classes with 2D data for image classification: Cats & Dogs, EyesPACS, Spyders & Chickens, and Shapes. After this testing, we evaluated DCON in 3D data for action recognition: UCF101 and Human2 (a dataset we compiled to control movement, clothing and background). The contribution of this research lies in a guideline to use DCON. It can be summarized as follows: if DCON is used on the first layers of the proposal of model (with simple features), the computational resources expressed as the quantity of Flops will tend to increase and produce bigger misclassification than the standard CNN. However, if the DCON is used at the end layers, the quantity of Flops used in the training and testing will decrease, and the classification accuracy will improve by up to 20% about the base model. Moreover, it gains robustness when using deformable convolutions because it can adapt to the region of interest. Also, the best kernel size of the DCON is three. It showed better results than size five. In the last case, the quantity of Flops increase quadratically, but their performance does not increase significantly. With these results, we propose a guideline to use the DCON and contribute to understanding the impact of DCON on the robustness of CNNs.

Many types of banana (Musa spp.)—one of the most economically important horticultural crops—exist, and their ploidy (diploid, triploid, and tetraploid, usually) and genome types (containing A or/and B genome in most cases) differ. At present, observation and genome type detection are commonly used to identify banana germplasm resources. However, the former is tedious, while the latter cannot distinguish categories below genome types. It is therefore urgent that a simple and effective method for identifying banana germplasm resources is established. We sequenced and analyzed the ribosomal DNA internal transcribed spacer (ITS) sequences of 62 banana germplasm and found that the sequencing peaks, especially the 20 bp region near the 420 bp position (referred to as the “420 bp region”), exhibited relatively recognizable and repeatable polymorphism characteristics. Using the “420 bp region” as a marker, we were able to quickly distinguish bananas belonging to different genome type groups (comprising the AA, AB, AAA, ABB, AAB, ABBB, AAAB, and AAAA groups) or different subgroups in the same genome type group (for example, the Cavendish, Gros Michel, Red, Lakatan, and Ibota Bota subgroups in the AAA group; and the Pisang Raja, Plantain, Pome and Silk subgroups in the AAB group). Moreover, it appeared that Sanger sequencing of ITS could be used for identifying banana hybrid offspring. These results demonstrated that the banana ITS region contains rich genetic information that is useful for genotyping. In general, ITS sequencing simplifies the classification of banana germplasm resources and has potential application in several areas of Musa improvement.

Text classification is one of the most important task in natural language processing, In this research, we carried out several experimental research on three (3) of the most popular Text classification NLP classifier in Convolutional Neural Network (CNN), Multinomial Naive Bayes (MNB), and Support Vector Machine (SVN). In the presence of enough training data, Deep Learning CNN works best in all parameters for evaluation with 77% accuracy, followed by SVM with accuracy of 76%, and multinomial Bayes with least performance of 69% accuracy. CNN has the best performance in the presence of large enough training dataset because of the presence of filter/ kernels which help to indentify patterns in text data regardless of their position in the sentence. We repeated the training again with just one-third of our data, at this point SVM comes with the best performance, the performance of CNN noticeably drops but still better than Multinomial Naive Bayes, the reason why SVM works best when we reduce the training data was because of its ability to look for a hyper-plane that creates a boundary between different classes of data so as to properly classify them, so we believed that getting the hyper-plane was more efficient when we reduce the dataset, hence reason for the good performance. Multinomial Naive Bayes have the least performance which we attributed to its assumption of independency between the features which sometimes does not hold true. We concluded that availability of data should be an important factor when choosing classifier for Natural Language Processing Text Classification task. CNN should be use in the presence of enough dataset, and SVM should be use when data is not enough. Multinomial Naive Bayes must not be trusted with state of the art NLP task due to its assumption of independency between the features.

This paper introduces a novel framework for zero-shot learning (ZSL), i.e., to recognize new categories that are unseen during training, by distilling knowledge from foundation models. Specifically, we first employ ChatGPT and DALL-E to synthesize reference images of unseen categories from text prompts. Then, the test image is aligned with text and reference images using CLIP and DINO. Finally, the predicted logits are aggregated according to their confidence to produce the final prediction.Experiments are conducted on multiple datasets, including CIFAR-10, CIFAR-100, and TinyImageNet. The results demonstrate that our model can significantly improve classification accuracy compared to previous approaches, achieving AUROC scores above 96\% across all test datasets. Our code is available at https://github.com/1134112149/MICW-ZIC.

This study aims to automate the optimization of a Full-Range speaker in an SUV's audio system according to the equal loudness contours principle. The input signal, and frequency responses of the amplifier and speaker were transferred to Matlab. Using ideal filter parameters, ten parametric equalizer models were created, and the speaker output was obtained using the convolution technique. The same filter settings were applied to the vehicle multimedia system, and experimental results were obtained. The simulation and experimental results were compared, showing high similarity with a Pearson correlation coefficient of 0.9295 and an RMSE of 2.29. Results were compared with the ideal contour. The Pearson correlation coefficients for simulation and experimental results were 0.6341 and 0.6715, with RMSE values of 4.88 and 2.57, showing low similarity. Consequently, each parametric equalizer filter's parameters were optimized using a genetic algorithm. The optimized parameters were applied to the vehicle multimedia system, and results were compared with the ideal contour. The Pearson correlation coefficients for simulation and experimental results were 0.965 and 0.9661, with RMSE values of 2.12 and 1.48. These results indicate that optimization aligns the speaker output closer to the ideal contour, enhancing in-vehicle audio system performance and increasing user satisfaction.

This study explores advanced feature engineering and semantic enrichment methods to enhance text classification, focusing on detecting figurative language in tweets. The novel features introduced, Syno\_Lower\_Mean and Syn\_Mean, measure the use of uncommon synonyms and the mean frequency of synonyms, capturing semantic richness crucial for detecting figurative expressions. Using resources like SenticNet and Framester, we enrich our feature set with sentiment and frame semantic information. Our approach includes extensive data preprocessing, sophisticated feature selection, and implementing various classification models, such as SVM, KNN, Logistic Regression, Decision Trees, Random Forest, BERT, and LSTM networks. We rigorously evaluate each model's performance to assess the effectiveness of our features and enrichment methods. Putting emphasis on model explainability, we use decision tree analysis, feature importance analysis, and the TREPAN algorithm to approximate SVM decisions. Although we focus on figurative language detection, our methods have broader implications for various NLP text classification tasks. Our findings demonstrate significant improvements in classification accuracy and interpretability through innovative feature design and dataset enrichment.

The automatic classification of bird sounds is an ongoing research topic and several results have been reported for the classification of selected bird species. In this contribution we use an artificial neural network fed with pre-computed sound features to study the robustness of bird sound classification. We investigate in detail if and how classification results are dependent on the number of species and the selection of species in the subsets presented to the classifier. In more detail, a bag-of-birds approach is employed to randomly create balanced subsets of sounds from different species for repeated classification runs. The number of species present in each subset is varied between 10 and 300, randomly drawing sounds of species from a dataset of 659 bird species taken from Xeno-Canto database. We observe that the shallow artificial neural network trained on pre-computed sound features is able to classify the bird sounds relatively well. The classification performance is evaluated using several common measures such as precision, recall, accuracy, mean average precision and area under the receiver operator characteristics curve. All of these measures indicate a decrease in classification success as the number of species present in the subsets is increased. We analyze this dependence in detail and compare the computed results to an analytic explanation assuming dependencies for an idealized perfect classifier. Moreover, we observe that the classification performance depends on the individual composition of the subset and varies across 20 randomly drawn subsets.

The detection and recognition of tables in image-based documents is a challenging and essential task for automated data extraction and anal- ysis. This paper introduces a methodology that utilizes Convolutional Neural Networks (CNNs) to address this problem. By harnessing the ro- bust visual recognition abilities of CNNs, the proposed method effectively detects and extracts tables from image-based documents. Additionally, an Encoder-Dual Decoder (EDD) architecture is employed to extract the textual content from table cells, enabling the transformation of visual data into structured, machine-readable formats. This structured data is then further processed using a Vector Space Model (VSM)-based language modeling technique for query-based table retrieval. The methodology’s accuracy is assessed by analyzing the spatial relationships between the extracted cells and comparing them against a pre-defined table structure. The results provide an evidence to the effectiveness of the proposed ap- proach, with an impressive 85% accuracy rate and a corresponding error rate of 15%. This accurate extraction and recognition of tables from image-based documents can significantly enhance applications in data analysis, information retrieval, and knowledge extraction.

The paper aims to develop and test an algorithm for classifying Russian-language text reviews of patients’ experiences with medical facilities and physicians, extracted from social media. This is motivated by the limitations of conventional methods of surveying consumers to assess their satisfaction with the quality of services, which are being replaced by automatic processing of text data from social media. This approach enables to get more objective results due to the increased representativeness and independence of the sample of service consumers. The authors have tested machine learning methods using various neural network architectures. A hybrid method was developed to classify text reviews of medical facilities posted by patients on the two most popular physician review websites in Russia. Overall, more than 60,000 reviews were analysed. The main results are as follows: 1) the classification algorithm developed by the authors has a high efficiency, the best result being achieved by the GRU-based architecture (val_accuracy = 0.9271); 2) applying the named entity search method to text messages following their partitioning improved the classification efficiency for each of the classifiers based on artificial neural networks. To further enhance the classification quality, reviews need to be semantically partitioned by target and sentiment and the resulting fragments need to be analysed separately.

Explosion monitoring is performed by infrasound and seismoacoustic sensor networks that are distributed globally, regionally, and locally. However, these networks are unevenly and sparsely distributed, especially in the local scale as maintaining and deploying networks is costly. With increasing interest in smaller yield explosions, the need for more dense networks has increased. To address this issue, we propose using smartphone sensors for explosion detection as they are cost-effective and easy to deploy. Although there are studies using smartphone sensors for explosion detection, the field is still in its infancy and new technologies need to be developed. We applied a machine learning model for explosion detection using smartphone microphones. The data used were from the Smartphone High-explosion Audio Recordings Dataset (SHAReD), a collection of 326 waveforms from 70 high-explosive (HE) events recorded on smartphones, and the ESC-50 dataset, a benchmarking dataset commonly used for environmental sound classification. Two machine learning models were trained and combined into an ensemble model for explosion detection. The resulting ensemble model classified audio signals as either “explosion,” “ambient,” or “other” with true positive rates (recall) greater than 96% for all three categories.

Hyperspectral images (HSI) contain abundant spectral information. Efficient extraction and utilization of this information for image classification remain prominent research topics. Previously, hyperspectral classification techniques primarily relied on statistical attributes and mathematical models of spectral data. Deep learning classification techniques have recently been extensively utilized for hyperspectral data classification, yielding promising outcomes. This study proposes a deep learning approach that uses polarization feature maps for classification. Initially, the polar coordinate transformation method is employed to convert the spectral information of all pixels in the image into spectral feature maps. Subsequently, the proposed Deep Context Feature Fusion Network (DCFF-NET) is utilized to classify these feature maps. The model is validated using three open-source hyperspectral datasets: Indian Pines, Pavia University, and Salinas. The experimental results indicated that DCFF-NET achieves excellent classification performance. Experimental results on three public HSI datasets demonstrate that the proposed method accurately recognizes different objects with an overall accuracy (OA) of 86.68%, 94.73%, and 95.14% based on the pixel method, and 98.15%, 99.86%, and 99.98% based on the pixel-patch method.

Incorrect sitting posture, characterized by asymmetrical or uneven positioning of the body, often leads to spinal misalignment and muscle tone imbalance. Prolonged maintenance of such postures can adversely impact well-being and contribute to the development of spinal deformities and musculoskeletal disorders. In response, smart-sensing chairs equipped with cutting-edge sensor technologies have been introduced as a viable solution for real-time detection, classification, and monitoring of sitting postures, aiming to mitigate the risk of musculoskeletal disorders and promote overall health. This comprehensive literature review evaluates the current body of research on smart-sensing chairs, with a specific focus on the strategies used for posture detection and classification, as well as the effectiveness of different sensor technologies. A meticulous search across MDPI, IEEE, and Google Scholar databases yielded 34 pertinent studies that utilize non-invasive methods for posture monitoring. The analysis reveals that Force Sensing Resistors (FSR) are the predominant sensors utilized for posture detection, whereas Convolutional Neural Networks (CNN) and Artificial Neural Networks (ANN) are the leading machine learning models for posture classification. However, it was observed that CNNs and ANNs do not outperform traditional statistical models in terms of classification accuracy, due to constrained size and lack of diversity within training datasets. These datasets often fail to comprehensively represent the array of human body shapes and musculoskeletal configurations. Moreover, this review identifies a significant gap in the evaluation of user feedback mechanisms, essential for alerting users to their sitting posture and facilitating corrective adjustments.

The use of artificial intelligence in natural language processing (NLP) has significantly contributed to the advancement of natural language applications such as sentimental analysis, topic modeling, text classification, chatbots, and spam filtering. With a large amount of text generated each day from different sources such as webpages, blogs, emails, social media, and articles, one of the most common tasks in natural language processing is the classification of a text corpus. This is important in many institutions for planning, decision-making, and archives of their projects. Many algorithms exist to automate text classification operations but the most intriguing of them is that which also learns these operations automatically. In this study, we present a new model to infer and learn from data using probabilistic logic and apply it to text classification. This model, called GenCo, is a multi-input single-output (MISO) learning model that uses a collaboration of partial classifications to generate the desired output. It provides a heterogeneity measure to explain its classification results and enables the reduction of the curse of dimensionality in text classification. The classification results are compared with those of conventional text classification models, and it shows that our proposed model has a higher classification performance than conventional models.

Domain generalization has led to remarkable achievements in Hyperspectral Image (HSI) classification. Inspired by contrastive language-image pre-training (CLIP), the language-aware domain generalization method has been explored to learn cross-domain-invariant representation. However, existing methods face some challenges: 1) The weak capacity to extract long-range contextual information and inter-class correlation. 2) Due to the inadequacies of the large-scale pre-training for HSI data, the spatial-spectral features of HSI and linguistic features can not be straightforwardly alignment. To address the above problems, a novel network has been proposed with a CLIP framework, which consists of an image encoder, based on an encoder-only transformer to obtain the global contextual information and inter-class correlation, a frozen text encoder, and a cross-attention mechanism, named Linguistic-Interact-with-Visual Engager (LIVE), enhances the interaction between two modalities. Extensive experiments demonstrating superior performance over state-of-the-art methods in HSI Domain Generalization with a CLIP framework.

Gastric cancer is the fifth most common and fourth deadliest cancer worldwide, with a bleak 5-year survival rate of about 20%. Despite significant research into its pathobiology, prognostic predictability remains insufficient due to pathologists’ heavy workloads and the potential for diagnostic errors. Consequently, there is a pressing need for automated and precise histopathological diagnostic tools. This study leverages Machine Learning and Deep Learning techniques to classify histopathological images into healthy and cancerous categories. By utilizing both handcrafted and deep features and shallow learning classifiers on the GasHisSDB dataset, we conduct a comparative analysis to identify the most effective combinations of features and classifiers for differentiating normal from abnormal histopathological images without employing fine-tuning strategies. Our methodology achieves an accuracy of 95% with the SVM classifier, underscoring the effectiveness of feature fusion strategies. Additionally, cross-magnification experiments produced promising results with accuracies close to 80% and 90% when testing the models on unseen testing images with different resolutions.

The ability to detect the presence as well as classify the activities of individuals behind visually obscuring structures is of significant benefit to police, security and emergency services in many situations. This paper presents the analysis from a series of experimental results generated using a through-the-wall (TTW) Frequency Modulated Continuous Wave (FMCW) C-Band radar system named Soprano. The objective of this analysis was to classify whether an individual was carrying an item in both hands or not using micro-Doppler information from a FMCW sensor. The radar was deployed at a standoff distance, of approximately 0.5 m, outside a residential building and used to detect multiple people walking within a room. Through the application of digital filtering, it was shown that significant suppression of the primary wall reflection is possible, significantly enhancing the target signal to clutter ratio. Singular Value Decomposition (SVD) signal processing techniques were then applied to the micro-Doppler signatures from different individuals. Features from the SVD information have been used to classify whether the person was carrying an item or walking free handed. Excellent performance of the classifier was achieved in this challenging scenario with accuracies up to 94%, suggesting that future through wall radar sensors may have the ability to reliably recognize many different types of activities in TTW scenarios using these techniques.

Background: Obesity remains a widely debated issue, often criticized for the limitations in its identification and classification. This study aims to compare two distinct systems for classifying obesity: Body Mass Index (BMI) and body fat percentage (BFP) as assessed by bioelectrical impedance analysis (BIA). By examining these measures, the study seeks to clarify how different metrics of body composition influence the identification of obesity-related risk factors. Methods: The study enrolled 1255 adults, comprising 471 males and 784 females, with a mean age of 36 ± 11.90 years. Participants exhibited varying degrees of weight status, including optimal weight, overweight, and obesity. Body composition analysis was conducted using the TANITA Body Composition Analyzer BC-418 MA III device (T5896, Tokyo, Japan), evaluating the following parameters: current weight, basal metabolic rate (BMR), adipose tissue (%), muscle mass (%), and hydration status (%). Results: Age and psychological factors like cravings, fatigue, stress, and compulsive eating were significant predictors of obesity in the BMI model but not in the BFP model. Additionally, having a family history of diabetes was protective in the BMI model (OR: 0.33) but increased risk in the BFP model (OR: 1.66). The BMI model demonstrates exceptional predictive ability (AUC=0.998). In contrast, the BFP model, while still performing well, exhibits a lower AUC (0.975), indicating slightly reduced discriminative power compared to the BMI model. Conclusions:BMI classification demonstrates superior predictive accuracy, specificity, and sensitivity. This suggests that BMI remains a more reliable measure for identifying obesity-related risk factors compared to the BFP model.

Cancer ranks second among the causes of mortality worldwide,following cardiovascular diseases.Brain cancer, in particular, has lowest survival rate of any kind of cancer. Brain tumours vary in their morphology, texture, and location, which determine their classification. Accurate diagnosis of the tumour category enables physicians to select optimal treatment strategies and potentially prolong patients’ lives. Researchers who also implemented deep learning models for diagnosing diseases in recent years largely focused on deep neural network optimization to enhance neural network performance. This implicates implementing models with the best performance and incorporating various network architectures by configuring their hyperparameters. This paper presents a novel hybrid approach method for improved brain tumour classification by combining CNN and EfficientNetV2B3 as feature extraction, followed by (KNN) for classification. To evaluate the recommended method’s efficacy, two widely known benchmark MRI datasets were utilized in the experiments. The initial dataset consisted of 3064 MRI images depicting meningiomas,pituitary,and gliomas tumours.Images from two classes of normal brain and brain tumour were included in the second dataset, which was obtained from Kaggle.In order to enhance performance even more, the study concatenates the CNN and EfficientNetV2B3 flattened outputs before feeding them into the KNN classifier. The proposed framework run on two different dataset and demonstrates outstanding performance with an accuracy of 99.51% and 99.8% on each dataset.

In this paper, we propose an imagined speech-based brain wave pattern recognition using deep learning. Multiple features were extracted concurrently from eight-channel Electroencephalography (EEG) signals. To obtain classifiable EEG data with fewer number of sensors, we placed the EEG sensors on carefully selected spots on the scalp. To decrease the dimensions and complexity of the EEG dataset and to avoid overfitting during the deep learning algorithm, we utilized the wavelet scattering transformation. A low-cost 8-channel EEG headset was used with MATLAB 2023a to acquire the EEG data. The Long-Short Term Memory Recurrent Neural Network (LSTM-RNN) was used to decode the identified EEG signals into four audio commands: Up, Down, Left, and Right. Wavelet scattering transformation was applied to extract the most stable features by passing the EEG dataset through a series of filtration processes. Filtration has been implemented for each individual command in the EEG datasets. The proposed imagined speech-based brain wave pattern recognition approach achieved a 92.50% overall classification accuracy. This accuracy is promising for designing a trustworthy imagined speech-based Brain-Computer Interface (BCI) future real-time systems. For better evaluation of the classification performance, other metrics were considered, and we obtained 92.74%, 92.50% and 92.62% for precision, recall, and F1-score, respectively.

Transmission of values and religious concepts to children is one of the most important issues in the third millennium and it has drawn varied and different views among experts and scholars in the world. Research specialized in religious literature for children and adolescents create new capacities in the presentation of religious concepts to the group. Plans have been considered to transfer values and religious concepts in the curricula of primary school children in the group in Iran. It is one of the topics that the authors note to the introduction of the minutiae of religion in the first three elementary grades. In this study, the collection and analysis methods providing content related to the minutiae of religion in reading books the first till third sections of the years 2013-2015. In addition, the plan includes aspects of other branches of religion in these books on information collected from text books that collected and classified. The result is that "definitely good and forbidding the evil" and "prayer" have the highest frequency of applications in the selected books. Further branches made of branches of religion in these books, represents the values of religious, moral and social as well.

In this paper, a seamless connection between sensor technology and machine learning (ML) has been presented which has wide applicability in rehabilitation and physical medicine. The intent is to identify an ML algorithm which has a better performance, easy to deploy, and has less training time and consumes less computational resources. In short, we propose an algorithm which maximizes the synergy between a sensor and ML by maximizing classification performance and easy to do implementation, which is robust in computational power and lower in resource consumption. The premise of the work is to show the important considerations for evaluating ML algorithms for Upper-limb activity recognition for Physical medicine and rehabilitation field.

This paper propose an original approach of achieving a Cymatics based visual perception of image-extracted text. In this context, an effective approach for automated text detection and recognition for the natural scene images is proposed. The incoming image is firstly enhanced by employing CLAHE and DWT. Afterwards, the text regions of the enhanced image are detected by employing the MSER feature detector. The non-text MSERs are removed by employing the geometrical and contour based filters. The remaining MSERs are grouped into words or phrases by finding out similarities between them. The text recognition is performed by employing an OCR function. The extracted text is sequentially analysed on character by character basis. Each character is converted into a methodical acoustic excitation. Finally, these excitations are converted into the systematic visual perceptions by using the phenomenon of Cymatics. The system functionality is tested with an experimental setup. For the case of studied natural scenes, the suggested approach achieves 80% precision in text localization and 53% precision in end-to-end text recognition. The devised system principle is novel and can be employed in various applications like visual art, encryption, education, integration of impaired people, etc.

Highly diverse agroecosystems are increasingly of interest as the realization of farms' invaluable ecosystem services grows. Simultaneously there has been an increased use of uncrewed aerial systems (UAS) in remote sensing as drones offer a finer spatial resolution and faster revisit rate than traditional satellites. With the combined utility of UAS and the attention on agroecosystems, there exists an opportunity to assess UAS practicality in highly biodiverse settings.  In this study, we utilized UAS to collect fine-resolution 10-band multispectral imagery of coffee agroecosystems in Puerto Rico. We created land cover maps through a pixel-based supervised classification of each farm and assembled accuracy assessments for each classification. To bolster our understanding of the classifications, we interviewed farmers to understand their thoughts on how these maps may be best used to support their land management. The average overall accuracy (53.9%), though relatively low, was expected for such a diverse landscape with fine-resolution data. After sharing imagery and land cover classifications with farmers, we found that while the prints were often a point of pride or curiosity for farmers, integrating the maps into farm management was perceived as impractical. These findings highlight that while remote sensing of diverse agroecosystems may provide a detailed way of estimating land cover classes and ecosystem services for researchers and government agencies for example these maps may be of limited use to land managers without additional interpretation.

The research presents the Methodology of Improving the Accuracy in Text Classification in Light of Modelling the Latent Semantic Relations (LSR). The aim of this Methodology is to find the ways of eliminating the Limitations of Discriminant and Probabilistic methods for LSR revealing and customizing the Text Classification Process to the more accurate recognition of the text tonality. This aim should be achieved by using the knowledge about the text’s Hierarchical Semantic Context in the form of Corpora-based Hierarchical Sentiment Dictionary. The main scientific contribution of this research is the following set of approaches to improve the qualitative characteristics of Text Classification process: combination of the Discriminant and Probabilistic methods allowing to decrease the influences of the Limitations of these methods on the LSR revealing process; considering each document as a complex structure allowing to estimate documents integrally by separated classification of topically completed textual component (paragraphs); taking into account the features of Argumentative type of documents (Reviews) allowing to use the author’s subjective evaluation of text tonality for development the Text Classification methodology. Tonality, expressed by the Review’s author, has a significant, but not critical, effect on the qualitative indicators of Sentiment Recognition.

Background: the primary care service in Catalonia has operated an asynchronous teleconsulting service between GPs and patients since 2015 (eConsulta), which has generated some 500,000 messages. New developments in big data analysis tools, particularly those involving natural language, can be used to accurately and systematically evaluate the impact of the service. Objective: the study was intended to examine the predictive potential of eConsulta messages through different combinations of vector representation of text and machine learning algorithms and to evaluate their performance. Methodology: 20 machine learning algorithms (based on 5 types of algorithms and 4 text representation techniques)were trained using a sample of 3,559 messages (169,102 words) corresponding to 2,268 teleconsultations (1.57 messages per teleconsultation) in order to predict the three variables of interest (avoiding the need for a face-to-face visit, increased demand and type of use of the teleconsultation). The performance of the various combinations was measured in terms of precision, sensitivity, F-value and the ROC curve. Results: the best-trained algorithms are generally effective, proving themselves to be more robust when approximating the two binary variables "avoiding the need of a face-to-face visit" and "increased demand" (precision = 0.98 and 0.97, respectively) rather than the variable "type of query"(precision = 0.48). Conclusion: to the best of our knowledge, this study is the first to investigate a machine learning strategy for text classification using primary care teleconsultation datasets. The study illustrates the possible capacities of text analysis using artificial intelligence. The development of a robust text classification tool could be feasible by validating it with more data, making it potentially more useful for decision support for health professionals.

In order to achieve the elimination of the negative impacts of transport on road safety, the European Union is taking various measures resulting from its commitment to improve road safety. One of the instruments for achieving the objectives set is social legislation for the work of drivers in road freight and passenger transport. In line with the objective of improving the quality of working conditions for road transport drivers and road safety, the European Union applies Regulation (EC) No 561/2006 on the harmonisation of certain social legislation relating to road transport, which lays down rules on driving times, breaks and rest periods for drivers operating in road freight and passenger transport. The main objective of the paper is to assess the impact of social legislation on road transport safety by means of real research and to identify the factors influencing the violation of the legislation depending on the type of transports carried out in domestic or international road freight transport.

Abstract This study delved into the application of various machine learning (ML) models for the classification of lung cancer levels. Through meticulous monitoring of parameters such as minimum child weight and learning rate, efforts were made to mitigate overfitting while optimizing model performance. The Deep Neural Network (DNN) emerged as a standout performer, showcasing robust performance across training, validation, and testing stages. Ensemble methods like voting and bagging also demonstrated promising results. However, Support Vector Machine (SVM) models with the Sigmoid kernel faced challenges in achieving satisfactory performance. Overall, the investigation sheds light on the efficacy of different ML models in lung cancer level classification and underscores the importance of parameter tuning to address overfitting concerns.

Detection of subsurface archaeological remains using a range of remote sensing methods, poses several challenges until today. Recent studies regarding the detection of archaeological proxies like those of cropmarks highlight the complexity of the phenomenon. In this work we present three different methods, and associated indices, for identifying stressed reflectance signatures indicating buried archaeological remains, based on a dataset of measured ground spectroradiometric reflectance. Several spectral profiles between the visible and near infrared part of the spectrum, were taken over a controlled environment in Cyprus during the 2011-2012 and are re-used in this study. The first two (spectral) methods are based on a suitable analysis of the spectral signatures in (1) the visible part of the spectrum, and in particular in the neighborhood of 570 nm, and (2) the red edge-near infrared part of the spectrum, in the neighborhood of 730 nm. Machine learning (decision trees) allows for the deduction of suitable wavelengths to focus on, in order to formulate the proposed indices and the associated classification criteria (decision boundaries) that can enhance the detection probability of stressed vegetation. Noise in the signal is taken into account by simulating reflectance signatures perturbed by white noise. Applying decision tree classification on the ensemble of simulations and basic statistical analysis we refine the formulation of the indices and criteria for the noisy signatures. The success rate of the proposed methods is over 90%. The third method rests on the estimation of vegetation/canopy reflectance parameters through inversion of the physical-based PROSAIL reflectance model and the associated classification through machine learning methods. The obtained results provide further insights into the formation of stress vegetation that occurred due to the presence of shallow buried archaeological remains, which are well aligned with physical-based models and existing empirical knowledge. To the best of the authors' knowledge, this is the first study demonstrating the usefulness of radiative transfer models such as PROSAIL for understanding the formation of cropmarks. Similar studies can support future research directions towards the development of regional remote sensing methods and algorithms if systematic observations are adequately dispersed in space and time.

The application of artificial intelligence to point-of-care testing (POCT) disease detection has become a hot research field, in which breath detection which detects the patient's exhaled VOCs combined with sensor arrays of Convolutional Neural Network (CNN) algorithms as a new lung cancer detection is attracting more researchers' attention. However, due to low accuracy, high complexity computation and large number of parameters make the CNN algorithms difficult transplant to the embedded system of POCT devices. A lightweight neural network (LTNet) in this work is proposed to deal with this problem, and meanwhile, achieve high-precision classification of acetone and ethanol gases, which are respiratory markers for lung cancer patients. Compared to currently popular lightweight CNN models such as EfficientNet, LTNet has fewer parameters (32K) and its training weight size is only 0.155 MB. LTNet achieved an overall classification accuracy of 99.06% and 99.14% in the own mixed gas dataset and the University of California (UCI) dataset, which are both higher than that of the five existing models’, and it also offers the shortest training (844.38s and 584.67s) and inference time (23s and 14s) in the same validation set. Compared to the existing CNN models, LTNet is more suitable for resource-limited POCT devices.

Timely and accurate information on tree species is crucial for the sustainable management of natural resources, forest inventory, biodiversity detection, and carbon stock calculation. The advancement of remote sensing technology and artificial intelligence has facilitated the acquisition and analysis of remote sensing data, resulting in more precise and effective classification of tree species. Multimodal remote sensing data and deep learning seem to be the current tree species classification research mainstream, whether or not. The current review on the remote sensing data and deep learning tree species classification methods perspectives to analyze the unimodal and multimodal remote sensing data and classification methods in this realm is missing. To bridge the gap, we search for major trends in the remote sensing data and tree species classification methods, provide a detailed overview of classic deep learning-based methods for tree species classification, and discuss the limitations.

This study describes a novel method for grading pathological sections of gliomas. Our own integrated hyperspectral imaging system was employed to characterize 270 bands of cancerous tissue samples from microarray slides of gliomas. These samples were then classified according to the guidelines developed by the World Health Organization, which define the subtypes and grades of diffuse gliomas. We explored a hyperspectral feature extraction model called SMLME-Resnet using microscopic hyperspectral images of brain gliomas of different malignancy grades. The model combines the channel attention mechanism and multi-scale image features to automatically learn the pathological organization of gliomas and obtain hierarchical feature representations, effectively removing the interference of redundant information. It also completes multi-modal, multi-scale spatial and spectral feature extraction to improve the automatic classification of glioma subtypes. The proposed classification method demonstrated high average classification accuracy (>97.3\%) and Kappa coefficient (0.954), indicating its effectiveness in improving the automatic classification of hyperspectral gliomas. The method is readily applicable in a wide range of clinical settings, offering valuable assistance in alleviating the workload of clinical pathologists. Furthermore, the study contributes to the development of more personalized and refined treatment plans, as well as subsequent follow-up and adjustment of treatment, by providing physicians with insights into the underlying pathological organization of gliomas.

Traffic Inspection (TI) work in this article is positioned as a specific module of road traffic with its primary function oriented towards monitoring and controlling safe traffic and the execution of significant events within a particular geographic area. Exploratory research on the significance of event execution in simple, complicated and complex traffic flow and process situations is related to the activities of monitoring and controlling functional states and performance of categorical variables, which include objects and locations of road infrastructure, communication infrastructure and networks of traffic inspection resources. It is emphasized that the words “work” and “traffic” have the semantic status as synonyms (in one world language), which is explained in the design of the Agent-based model of the complexity of content and contextual structure of TI work at the singular and plural level with 12 points of interest (POI) in the thematic research. An Event Execution Log (EEL) was created for on-site data collection with eight variables, seven of which are independent (event type, activities, objects, locations, host, duration-period and periodicity of the event) and one dependent (significance of the event) variable. The structured data set includes 10,994 input-output vectors in 970 categories collected in the EEL created by 32 human agents (traffic inspectors) over a 30-day period. An algorithmic presentation of the methodological research procedure for preprocessing and final data processing in the ensemble of machine learning models for classification and selection of TI tasks is given. Data cleaning was performed on the available dataset to increase data consistency for further processing. Vector elimination was carried out based on the Location variable, such that the total number of vectors equals the number of unique categories of this variable, which is 636. The main result of this research is the Classification modeling of the significance of events in TI work based on machine learning techniques and the Stacking ensemble. The created machine learning models for event significance classification modeling have high accuracy values. To evaluate the performance metrics of the Stacking ensemble of models, the confusion matrix, Precision, Recall, ND F1-score are used.

The risk classification of barrier lake is the key to conducting emergency treatment in a scientific manner. Such difficulties are faced as short time for risk evaluation, complex evaluation indicators, difficulty in obtaining information quickly and quantifying index weights. Based on this, this paper constructs a quantitative risk classification model for barrier lakes based on D-AHP. On the basis of studies on nearly 100 cases of barrier lakes, an 8-factor evaluation index system and quantitative classification are proposed. The methods of rapid calculation of reservoir capacity curve of barrier lake and intelligent identification of particles on the surface of barrier body were developed, which realized the rapid acquisition of 8-factor evaluation index information in emergency environment. The D-AHP method dealt with inconsistent weight assignment to evaluation factors by experts, which helped achieve weight quantification of 8 factors. The risk assessment on15 barrier lakes such as Tangjiashan barrier lake show that the conclusion drawn with risk classification method proposed in this paper is basically consistent with that of the traditional table-lookup method. However, the table-lookup method ignores cumulative loss’s impacts on the risk level of barrier lakes, and considers the extremely severe loss of barrier lakes as a sufficient condition for the evaluation level to be grade I, and thus a deviation in the evaluation. The risk classification method proposed in this paper is more reasonable and reliable.

Surface water distribution extracted from remote sensing data has been used in water resource assessment, coastal management, and environmental change studies. Traditional manual methods for extracting water bodies cannot satisfy the requirements for mass processing of remote sensing data; therefore, accurate automated extraction of such water bodies has remained a challenge. The histogram bimodal method (HBM) is a frequently used objective tool for threshold selection in image segmentation. The threshold is determined by seeking twin peaks, and the valley values between them; however, automatically calculating the threshold is difficult because complex surfaces and image noise which lead to not perfect twin peaks (single or multiple peaks). We developed an operational automated water extraction method, the modified histogram bimodal method (MHBM). The MHBM defines the threshold range of water extraction through mass static data; therefore, it does not require the identification of twin histogram peaks. It then seeks the minimum values in the threshold range to achieve automated threshold. We calibrated the MHBM for many lakes in China using Landsat 8 Operational Land Imager (OLI) images, for which the relative error (RE) and squared correlation coefficient (R2) for threshold accuracy were found to be 2.1% and 0.96, respectively. The RE and root-mean-square error (RMSE) for the area accuracy of MHBM were 0.59% and 7.4 km2. The results show that the MHBM could easily be applied to mass time-series remote sensing data to calculate water thresholds within water index images and successfully extract the spatial distribution of large water bodies automatically.

This study examines the feasibility of using event-related potentials (ERPs) obtained from electroencephalographic (EEG) recordings as biomarkers for long-term memory item classification. Previous studies have identified old/new effects in memory paradigms associated with explicit long-term memory and familiarity. Recent advancements in convolutional neural networks (CNNs) have enabled the classification of ERP trials under different conditions and identification of features related to neural processes at the single-trial level. We employed this approach to compare three CNN models with distinct architectures using experimental data. Participants (N = 25) performed an association memory task while recording ERPs that were used for training and validation of the CNN models. The EEGNET-based model achieved the most reliable performance in terms of precision, recall, and specificity compared with the shallow and deep convolutional approaches. The classification accuracy of this model reached 70% for known items and 62% for unknown items. Good overall accuracy requires a trade-off between recall and specificity and depends on the architecture of the model and the dataset size. These results suggest the possibility of integrating ERP and CNN into online learning tools and identifying the underlying processes related to long-term memorization.

The land use land cover (LULC) map is extensively employed for different purposes. The LULC maps provided by the Portuguese National Geographic Information System (SNIG) are not ideal for continuous assessment and regular scrutiny. Machine learning (ML) algorithms applied in remote sensing (RS) data has been proven effective in image classification, object detection, and semantic segmentation. Previous studies have shown that Random Forest (RF) and Support Vector Machine (SVM) consistently achieved high accuracy for land classification. In this study, the application of RF and SVM classifiers, and object-based (OBIA) and pixel-based (PBIA) approaches with Sentinel-2A imagery was evaluated using Google Earth Engine (GEE) platform for land cover classification of a burned area in the Serra da Estrela Natural Park (PNSE), Portugal. This aimed to detect the land cover change and closely observe the burned area and vegetation recovery after the 2022 wildfire. The combination of RF and OBIA achieved the highest accuracy in all evaluation metrics. At the same time, a comparison with the Normalized Difference Vegetation Index (NDVI) map and Conjunctural Land Occupation Map (COSc) of 2023 year indicated that the SVM and PBIA map resembled the maps better.

This paper focuses on optimal sensor positioning for monitoring activities of daily living and investigates different combinations of features and models on different sensor positions, i.e., the side of the waist, front of the waist, chest, thigh, head, upper arm, wrist, and ankle. Sixteen features are extracted and the feature importance is measured by using the Relief-F feature selection algorithm. Eight classification algorithms are evaluated on a dataset collected from young subjects and that collected from elderly subjects, with two different experimental settings. To deal with different sampling rates, signals with a high data rate are down-sampled and a transformation matrix is used for aligning signals to the same coordinate system. The thigh, chest, side of the waist, and front of the waist are the best four sensor positions for the first dataset (young subjects), with average accuracy values being greater than 95%. The best model obtained from the first dataset for the side of the waist is validated on the second dataset (elderly subjects). The most appropriate number of features for each sensor position is reported. The results provide a reference for building activity recognition models for different sensor positions, as well as for data acquired from different hardware platforms and subject groups.

Identifying punch types as part of punch kinematic analysis is a key component for both coaches and players, providing critical insights into the variety and effectiveness of punches, which are essential for refining overall performance. Existing research indicates that punch type classification typically relies on data from either wearables or video data. Both approaches have their merits and demerits but require a significant amount of labeled data for modeling the classification algorithm. In this work, we propose a novel approach that significantly reduces the labeling effort by one-sixth by leveraging video data to enhance the training information of sensor data. This enriched training phase enables more accurate punch classification with reduced labeled data. Subsequently, the real-time punch identification and classification use only the wearable data. On a real dataset, with 15% of the labeling efforts, our approach achieved an accuracy rate of 91.41% for rear hand punch recognition and 91.91% for lead hand punch recognition, along with 92.33% and 94.56% for punch classification, respectively. Through the development of our Smart Boxer system, we aim to revolutionize punch analytics in boxing, providing comprehensive insights while enhancing the overall training and competitive experience for athletes and enriching fan engagement with the sport.

Aerial image classification has become popular and has attracted extensive research efforts in recent decades. The main challenge lies in its very high spatial resolution but relatively insufficient spectral information. To this end, spatial-spectral feature extraction is a popular strategy for classification. However, parameter determination for that feature extraction is usually time-consuming and depends excessively on experience. In this paper, an automatic spatial feature extraction approach based on image raster and segmental vector data cross-analysis is proposed for the classification of very high spatial resolution (VHSR) aerial imagery. First, multi-resolution segmentation is used to generate strongly homogeneous image objects and extract corresponding vectors. Then, to automatically explore the region of a ground target, two rules, which are derived from Tobler’s First Law of Geography (TFL) and a topological relationship of vector data, are integrated to constrain the extension of a region around a central object. Third, the shape and size of the extended region are described. A final classification map is achieved through a supervised classifier using shape, size, and spectral features. Experiments on three real aerial images of VHSR (0.1 to 0.32 m) are done to evaluate effectiveness and robustness of the proposed approach. Comparisons to state-of-the-art methods demonstrate the superiority of the proposed method in VHSR image classification.

Lung cancer remains a pervasive global health challenge, necessitating precise and efficient classification methods to inform optimal treatment strategies. In this study, we investigate the potential of machine learning algorithms in developing a computer-aided diagnostic tool for early and accurate lung cancer classification, leveraging essential protein biomarkers as predictive features. Our investigation reveals the superiority of Deep Neural Networks (DNNs) in lung cancer classification, achieving an impressive accuracy of 96.91%. This highlights the transformative potential of DNNs in facilitating real-world clinical applications, offering clinicians a powerful tool for accurate diagnosis and prognosis. Additionally, we analyze the performance of Ensemble Methods, including Voting (91.75%) and Bagging (93.81%), as viable alternatives to DNNs. These ensemble techniques demonstrate robust performance, further underlining their utility in lung cancer classification tasks. Furthermore, our study underscores the critical role of hyperparameter tuning, particularly in adjusting parameters such as min child weights and learning rates, in mitigating overfitting and enhancing the generalizability of diverse machine learning models. For instance, Support Vector Machines (SVM) exhibit varying accuracies ranging from 74.23% to 95.88% across different hyperparameter configurations, emphasizing the importance of hyperparameter optimization in maximizing model performance. In summary, our comprehensive comparative analysis sheds light on the efficacy of different machine learning approaches in lung cancer classification. By leveraging advanced algorithms and optimizing model parameters, researchers and clinicians can harness the power of machine learning to advance early detection and personalized treatment strategies for lung cancer patients.

ACL instability poses a significant challenge in traumatology and orthopedic medicine, often requiring accurate diagnosis for appropriate treatment. While the pivot-shift test offers a crucial means of assessment, its reliance on subjective interpretation underscores the need for supplementary imaging studies. This study aims to address this limitation by developing a Bayesian classification algorithm tailored for integration into a mobile application. Using the built-in inertial sensors of smartphones, this new approach aims to dynamically evaluate rotational stability during knee examinations. Orthopedic specialists conducted knee evaluations on 52 subjects, with subsequent analysis revealing interesting insights. Intraobserver and interobserver analyses, as measured by ICC, demonstrated strong agreement both in terms of timing between maneuvers (ICC = 0.94) and signal amplitude (ICC = 0.71-0.66). Notably, the Bayesian algorithm successfully classified 95% of joint hypermobility cases, with an additional 7 cases of hyperlaxity identified by the Pivot-Shift Meter (PSM). These findings highlight the practicality and effectiveness of implementing a Bayesian classification algorithm within a mobile application for assessing and categorizing signals captured by smartphone inertial sensors during the pivot-shift test.

In 2020, students of Pakistan had to adapt to the online environment for the very first time. This study aims to analyze the perceptions of medical, dental, and allied health students about online education in Pakistan. A descriptive, cross-sectional study was done to assess the level of acceptance of undergraduate students. A pre-validated questionnaire regarding demographics, past-experience of e-learning, advantages disadvantages of e-learning, and general perception of students towards e-learning was distributed. Descriptive statistics were computed for demographics, Mann-Witney-U test was used to compare the differences of perceptions between pre-clinical year and clinical years students. Kruskal-Wallis test was applied to compare the results of three specialties of students. Chi-square was used to compare overall category-wise positive and negative responses of students. 1200 students participated in the study. The major advantage identified by all students was the ‘comfortable environment’ in which they studied online. The major disadvantage selected by preclinical year students was ‘anxiety due to social isolation’ and that chosen by clinical year students was ‘lack of patient interaction’. Overall, 72% of students had a negative perception of e-learning. Student-teacher training, student counselling sessions, and innovative techniques need to be introduced to enhance student engagement and reduce pandemic stress.

Aiming at solving difficulties related to aero-engine classification and identification, the infrared spectra of aero-engine hot jets are measured by telemetry Fourier transform infrared spectrometer, and a six types aero-engine hot jet spectrum data set is generated. The spectrum data set were randomly sampled at a ratio of 8:1:1 to generate training set, validation set and prediction set. In this paper, a Peak Finding Siamese Convolutional Neural Network (PF-SCNN) is designed to match and classify spectrum data. In the first place, the matching pairs of training set and validation set are made independently with the labels of positive and negative samples creating. The spectrum feature extraction and distance similarity calculation are carried out by the Siamese convolutional neural network (SCNN). When entering the prediction phase, the prediction set and training set are integrated to predict in the trained model. The prediction set labels are ultimately determined by matching them with the training set labels having the highest similarities to ensure accurate predictions. To improve the operation efficiency of the SCNN, a peak finding method is introduced to extract the spectrum peaks. Peak positions of high frequency are counted, while the peak data are extracted. The peak data are used to take the place of the original data for training and prediction. The performance measures are defined as Accuracy, Precision, Recall, Confusion matrix and F1-score, and the prediction accuracy was up to 99%. The experimental results indicate that the proposed PF-SCNN is suitable for our spectral dataset and can complete the task of infrared spectrum classification of aero-engine hot jets.

Text mining is a technological trend often highlighted in the continuous exchange of information through interconnected media. Its applicability goes beyond private organizations, as the public sector also requires it to treat textual information regarding services offered. Within this scenario, public security emerges as a prominent user of text mining that seeks to ensure the construction of data and knowledgebases to support decision-making about law enforcement actions to ensure citizen welfare. The primary objectives of this article are: (i) to develop a survey to identify text mining applications, techniques, opportunities, and challenges in public security, and (ii) to outline research directions concerning these topics and provide insights so that interested researchers can develop new studies. The literature was searched within four databases: Scopus, IEEE Xplore, ACM Digital Library, and Web of Science. A filtering process was applied to extract the works most aligned with the target theme, resulting in the selection of 194 of the most relevant works for a literature review. There were identified nineteen key applications of text mining related to public security and the most recurrent techniques and technologies reported between 2014 to 2021, supporting outlining three axes for future directions: one with possible expansion of objectives for new research; another on changes and adaptations in scopes for the methodological context; and the last one on expansions and changes in application scenarios based on the literature.

The Upper Hunter Valley is a major coal mining area in New South Wales (NSW), Australia. Due to the ongoing increase of mining activities, PM10 (air-borne particle with an aerodynamic diameter less than 10 micrometres) pollution has become a major air quality concern in local communities. The early part of this study identified two air quality subregions in the valley. This paper aims to provide a holistic summarisation of the relationships between elevated PM10 pollution in two subregions and the local- and synoptic-scale meteorological conditions for spring and summer, when PM10 pollution is relatively high. A catalogue of twelve synoptic types and a set of six local meteorological patterns have been quantitatively derived and linked to each other using the self-organising map (SOM) technique. The complex meteorology-air pollution relationships have been visualised/interpretated on the SOM planes for two representative locations. It was found that the influence of local meteorological patterns differed significantly for mean PM10 levels vs the occurrence of elevated pollution events and between air quality subregions. In contrast, synoptic types showed generally similar relationships with mean vs elevated PM10 pollution in the valley. Two local meteorological patterns, the hot-dry-northwesterly wind conditions and the hot-dry-calm conditions, were found most PM10 pollution conducive in the valley when combined with a set of synoptic counterparts. These synoptic types are featured with the influence of eastward migrating continental high pressure system and westerly troughs, or a ridge extending northwest towards coastal northern NSW or southern Queensland from the Tasman Sea. The method and results can be used in air quality research for other locations of NSW, or similar regions elsewhere.

Timely and accurate acquisition of information on distribution of the crop planting structure in Loess Plateau of eastern Gansu Province, one of the most important agricultural areas in western China, is crucial for promoting fine management of agriculture and ensuring food security. In this study, the Object-Based Image Classification (OBIC), the Random Forest (RF) and Convolutional Neural Network (CNN) models were employed to classify the crop planting structure of four representative test areas in Qingyang City in a precise manner. Firstly, different optimal segmentation scales for various crops were selected using the Estimation of Scale Parameter 2 (ESP2) tool and the Ratio of Mean Difference to Neighbors(ABS) to Standard Deviation (RMAS)model. The images were then segmented through multiresolution segmentation combined with the Canny Edge Detection algorithm. Secondly, the L1 regularized logistic regression model was utilized to select and optimize 39 spatial feature factors including spectral, textural, geometric, and index features, in conjunction with phenological factors. Lastly, under the multi-level classification framework, the Random Forest (RF) classifier and Convolutional Neural Network (CNN) model, combined with object-based multiresolution segmentation, was used to classify the crop planting structure. The findings show that: Thanks to the Canny Edge Detection algorithm, we can obtain a more complete boundary of the segmented objects and improve the separability. The optimal segmentation scales for corn, vegetables, and buckwheat were found to be 55, 70, and 35, respectively, while wheat and apple had optimal segmentation scales of 65. In addition to phenological characteristics, the number of selected spatial features for corn, vegetables, buckwheat, wheat, and apple were 9, 7, 16, 12, and 10, respectively. The CNN model demonstrated high consistency with the RF model in the classification results, but the accuracy of RF model is higher than that of CNN model on the whole. The overall accuracy of classification using RF model in four test areas registered 91.93%, 94.92%, 89.37% and 90.68%, respectively. This paper introduced crop phenological factors, effectively improving the extraction precision of shattered agricultural planting structure in the Loess Plateau of eastern Gansu Province. Its findings have important application value in crop monitoring, management, food security and other related fields.

The advent of deep learning (DL) has revolutionized medical imaging, offering unprecedented avenues for accurate disease classification and diagnosis. DL models have shown remarkable promise for classifying brain tumor from Magnetic Resonance Imaging (MRI) scans. However, despite their impressive performance, the opaque nature of DL models poses challenges in understanding their decision-making processes, particularly crucial in medical contexts where interpretability is essential. This paper explores the intersection of medical image analysis and DL interpretability, aiming to elucidate the decision-making rationale of DL models in brain tumor classification. Leveraging state-of-the-art DL frameworks with transfer learning, we conduct a comprehensive evaluation encompassing both classification accuracy and interpretability. Using state-of-the-art DL frameworks with transfer learning, we conduct a thorough evaluation covering both classification accuracy and interpretability. We employ adaptive path-based techniques to understand the underlying decision-making mechanisms of these models. Grad-CAM and Grad-CAM++ highlight critical image regions where the tumors are located.

Stunting is a major public health issue in Aceh, Indonesia, requiring advanced analytical techniques for effective interventions. This study presents a novel hybrid machine learning framework designed to enhance the analysis of stunting through improved classification, predic-tive modeling, and clustering optimization. The framework utilizes Support Vector Machines (SVM) with Radial Basis Function (RBF) and Sigmoid kernels for classification. The RBF kernel achieved an accuracy of 91.3%, significantly outperforming the Sigmoid kernel's 85.6%. Linear Regression was employed for predictive modeling, yielding a Mean Squared Error (MSE) of 0.137, which indicates strong predictive accuracy. In clustering, the optimized K-Medoids method, in-corporating a weight product approach, demonstrated superior efficiency by requiring only 3 iterations for convergence, compared to 7 iterations for the conventional K-Medoids method. Additionally, it achieved a higher Calinski Harabasz Index of 93.7, compared to 85.2 for the conventional method. This comprehensive approach enhances accuracy and efficiency across classification, prediction, and clustering tasks, providing valuable insights for targeted interventions and policy development to address stunting in Aceh.

Electronic health records (EHRs) are becoming a vital source of data for healthcare quality improvement, research, and operations. However, much of the most valuable information contained in EHRs remains buried in unstructured text. The field of clinical text mining has advanced rapidly in recent years, transitioning from rule-based approaches to machine learning and, more recently, deep learning. With new methods come new challenges, however, especially for those new to the field. This review provides an overview of clinical text mining for those who are encountering it for the first time (e.g. physician researchers, operational analytics teams, machine learning scientists from other domains). While not a comprehensive survey, it describes the state of the art, with a particular focus on new tasks and methods developed over the past few years. It also identifies key barriers between these remarkable technical advances and the practical realities of implementation at health systems and in industry.

The primary objective of this study was to identify a rapid and noninvasive methodology for effectively differentiating various herbal species cultivated in Eastern Poland. Typically, a plant's chemical profile is obtained using chromatographic methods; however, spectroscopic methods can serve as a complementary or independent tool for determining similarities in the analyzed samples. The investigation centered on leaves of plant species from the Lamiaceae family, encompassing basil, lavender, oregano, sage, and thyme. UV-VIS and ATR-FTIR spectra of dried plant samples' powder were recorded and subsequently analyzed using chemometric tools. This paper assesses the relative performance of different discrimination methods, including both traditional methods based on machine learning and those dealing with specimens associated with a specific category. Among the methodologies tested, SIMCA showed the best performance, particularly effective for small training sets.

Cardiopulmonary resuscitation (CPR) is a widely practiced lifesaving skill. Conventional CPR is based on manual chest compressions and delivering mouth-to-mouth breaths. There has been advancement in biomedical technology over the last few decades which has led to the introduction of new techniques and devices in CPR as well. In light of these advancements, we propose SMEAR (Standard, Mechanical, ECMO Assisted, Augmented virtual and Remote) classification system to provide structured overview from traditional to novel methods and devices. The standard or Manual CPR, widely practiced by the general public during out-of-hospital cardiac arrests, combines active chest compressions with or without mouth breathing. However, to enhance CPR’s efficacy, various mechanical devices have been developed, encompassing Active Chest Compression, Passive Decompression, and those integrated with Automated External Defibrillators. Feedback sensor-equipped devices, include the Metronome, Accelerometer, Barometer, and Triaxial Field Induction. Mechanical devices, classified by their compression mechanisms, can be isolated – like the Piston-driven, Load-distributing, and Active Compression-Decompression (ACD) – or combined, as in ACD-ITV (Impedance Threshold Device), ACD-IAC (Interposed Abdominal Compression), and ACD-ITV-EECP (Enhanced External Counter Pulsation). Beyond mechanics, Extracorporeal CPR offers veno-arterial Extracorporeal Membrane Oxygenation in refractory cardiac arrest and can improve outcomes. Augmented Virtual Reality usage in training and provision of CPR is at the early stages. It has the potential of expanded usage in training public to provide high-quality CPR. A provider can be remotely guided step by step via telephone or smartphone applications, ensuring continuous support until professional assistance arrives.

Deep Learning (DL) oriented document processing is widely used in different fields for extraction, recognition, and classification processes from raw corpora of data. The article examines the application of deep learning approaches, based on different neural network methods, including Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), and Convolutional Neural Networks (CNN). The compared models were combined with two different word embedding techniques, namely: Bidirectional Encoder Representations from Transformers BERT and Gensim Word2Vec. The models are designed to evaluate the performance of architectures based on neural network techniques for the classification of CVs of Moroccan engineering students at ENSAK(National School of Applied Sciences of Kenitra, Ibn Tofail University). The used dataset included resumes collected from engineering students at ENSAK in 2023 for a project on the employability of Moroccan engineers in which new approaches were applied, especially machine learning, deep learning, and big data. Accordingly, 867 resumes were collected from five specialties of study (Electrical Engineering, Networks and Systems Telecommunications, Computer Engineering, Automotive Mechatronics Engineering, Industrial Engineering). The results revealed good performance of the proposed models based on the BERT embedding approach compared to models based on the Gensim Word2Vec embedding approach. Accordingly, the CNN-GRU/Bert model achieved slightly better accuracy with 0.9251 compared to other hybrid models.

Characters are one of the most important elements in composing digital animation. The appear-ance and voice of a character should be designed to express the personality and values of the character. However, it is not easy for animation producers to harmoniously match the appear-ance and voice of a character. Advances in deep learning technology have made it possible to overcome this limitation. To achieve this, firstly, an audio-visual dataset of characters is required. In this study, we construct and verify a Korean audio-visual dataset consisting of frontal face im-ages of various characters and short voice clips. We developed an application that can automati-cally extract the frontal face image and a short voice clip of a character by collecting videos up-loaded to YouTube. Through this, a dataset consisting of a total of 1,522 face images and a total of 7,999 seconds of voice clips was built based on 490 characters. Furthermore, we automatically la-bel characters by gender and age to validate the dataset. The dataset built in this study is expected to be used in various deep learning fields, such as classification, generative adversarial networks, and speech synthesis.

This article presents an overview of the first AI-human collaborated album, Hello World, by SKYGGE, which utilizes Sony’s Flow Machines technologies. This case study is situated within an overview of current and emerging uses of AI in popular music production, and connects those uses with myths and fears that have circulated in discourses concerning the use of AI in general, but also in music specifically. By proposing the concept of an audio uncanny valley in relation to AIPM (artificial intelligence popular music), this article offers a lens through which to examine the more novel and unusual melodies and harmonisations made possible through AI music generation, and questions how this content relates to wider speculations about posthumanism, sincerity, and authenticity in both popular music, and more wider assumptions of anthropocentric creativity. In its documentation of the emergence of a new era of popular music, the AI era, this article surveys: (1) the current landscape of artificial intelligence popular music focusing on the use of Markov models for generative purposes; (2) posthumanist creativity and the potential for an audio uncanny valley; and (3) issues of perceived authenticity in the technologically mediated “voice”.

In Indonesia, types of meat are still identified manually due to increasing prices for beef and buf-falo ahead of Idul Fitri. Traders mix beef with pork to thwart the beef scam. Haar wavelet and GLCM, GLCM with angles of 0°, 45°, 90°, and 135°, as well as matrices using contrast, correlation, energy, homogeneity, and entropy, are used in feature extraction. The following are the results of meat image classification testing using k-NN, Haar wavelet, and GLCM: The k-NN algorithm shows superiority in identifying fresh (99%), frozen (99%), and rotten (96%) meat texture images with the most fantastic accuracy results in every situation. Meanwhile, GLCM routinely provides good results, especially regarding the texture image of fresh meat (183.21) and rotten meat (115.79). However, despite delivering less throughput than k-NN and GLCM, Haar wavelets are still helpful, especially when dealing with fresh meat texture images (89,96).

In theory and applications, it is often inevitable to work with projectors onto convex sets, where a linear transform is involved. In this article, a novel projector is presented, which generalizes previous results in that it admits a broader family of linear transforms, but on the other hand it is limited to box-type convex sets in the transformed domain. The new projector has an explicit formula and it can be interpreted within the framework of proximal optimization. The benefit of the new projector is demonstrated on an example from signal processing, where it was possible to speed up the convergence of a signal declipping algorithm by a factor of more than two.

Transformer models have emerged as frontrunners in the field of natural language processing, primarily due to their adept use of self-attention mechanisms to grasp the semantic linkages between words in sequences. Despite their strengths, these models often face challenges in single-task learning scenarios, particularly when it comes to delivering top-notch performance and crafting strong latent feature representations. This challenge is more pronounced in the context of smaller datasets and is particularly acute for under-resourced languages such as Arabic. In light of these challenges, this study introduces a novel methodology for text classification of Arabic texts. This method harnesses the newly developed Reverse Positional Encoding (RPE) technique. It adopts inductive-transfer learning (ITL) framework combined with a switching self-attention shared encoder, thereby increasing the model's adaptability and improving its sentence representation accuracy. The integration of Mixture of Experts (MoE) and Reverse Positional Encoding (RPE) techniques empowers the model to process longer sequences more effectively. This enhancement is notably beneficial for Arabic text classification, adeptly supporting both the intricate five-point and the simpler ternary classification tasks. The empirical evidence points to its outstanding performance, achieving accuracy rates of 87.20% for the HARD dataset, 72.17% for the BRAD dataset, and 86.89% for the LABR dataset, as evidenced by the assessments conducted on these datasets.

Focusing on the problem of identifying and classifying aero-engine models, this paper measures the infrared spectrum data of aero-engine hot jets using a telemetry Fourier transform infrared spectrometer. Simultaneously, infrared spectral data sets with the six different types of aero-engines are created. For the purpose of classifying and identifying infrared spectral data, a CNN architecture based on the continuous wavelet transform peak seeking attention mechanism (CWT-AM-CNN) is suggested. This method calculates the peak value of middle wave band by continuous wavelet transform, and the peak data is extracted by the statistics of the wave number locations with high frequency. Attention mechanism is used for the peak data, and the attention mechanism is weighted to the feature map of the feature extraction block. The training set, vali-dation set and prediction set are divided in the ratio of 8:1:1 for the infrared spectral data sets. For three different data sets, CWT-AM-CNN proposed in this paper is compared with the classical classifier algorithm based on CO2 feature vector and the popular AE, RNN and LSTM spectral processing networks. The prediction accuracy of the proposed algorithm in the three data sets is as high as 97%, and the lightweight network structure design not only guarantees high precision, but also has a fast running speed, which can realize the rapid and high-precision classification of the infrared spectral data of the aero-engine hot jets.

The need for cross-modal retrieval (CMR) among users is growing because of the swift advancements in computer vision and natural language processing. Promoting home operating systems is greatly aided by the openKylin home operating system. It is focused on enhancing the functionality of the operating system with safe and dependable operating technologies. Practical implications abound for integrating computer vision research into home operating systems. A multimodal retrieval approach based on openKylin is presented to improve the usability of text search graphs, audio, and video. The overview of pertinent domestic and international research is the main topic of this article. This technological advancement. It then offers details on the design, implementation procedure, and final implementation outcomes for demonstration. The works of natural operations show that this strategy has excellent accuracy and performance.

Accurately predicting drug-drug interactions (DDIs) is crucial for preventing adverse drug events in clinical settings. However, existing methodologies often rely on complex models built from diverse data sources, posing challenges in computational drug discovery. To address the need for precise computational methods in predicting unknown DDIs, this study introduces a novel Deep Neural Network (DNN)-based approach. By leveraging a wide range of drug-related information, including substructure, targets, side effects, pathways, and indications, our method calculates multiple drug similarities. These similarities are then synthesized using a nonlinear fusion method to extract high-level features. Subsequently, a tailored neural network is deployed for interaction prediction. Comparative evaluation against three prominent machine learning classifiers—Extreme Gradient Boosting (XGBoost), Adaptive Gradient Boosting (AdaBoost), and Light Gradient-Boosting Machine (LGBM)—using three benchmark datasets demonstrates the superior performance of DNN. It achieves outstanding accuracy, precision, recall, and F1-score metrics, all reaching 94.6% in cross-validation. Additionally, case studies involving numerous drug pairs confirm the reliability of DNN in accurately predicting unknown DDIs. These findings underscore DNN as a potent and reliable method for DDI prediction, with promising implications for drug discovery and healthcare applications.

Information dissemination and preservation are crucial for societal progress, especially in the technological age. While technology fosters knowledge sharing, it also risks spreading misinformation. Audio deep fakes, convincingly fabricated audio using Artificial intelligence (AI), exacerbate this issue. We present Sonic Sleuth, an AI model for detecting audio deepfakes. Our approach leverages Deep Learning (DL) for a robust detection model. Meticulous data preprocessing and rigorous experimentation with various models led to the implementation of the most effective solution with a custom CNN model. Our comprehensive testing resulted in a highly accurate model (98.27% accuracy, 0.016 EER) trained on a substantial dataset of real and synthetic audio. In addition to an 84.92% accuracy and 0.085 EER on an external dataset, these results demonstrate Sonic Sleuth’s potential as a powerful tool against audio misinformation.

Polarimetric Synthetic Aperture Radar (PolSAR) images encompass valuable information that can facilitate extensive land cover interpretation and generate diverse output products. Extracting meaningful features from PolSAR data poses challenges distinct from those encountered in optical imagery. Deep Learning (DL) methods offer effective solutions for overcoming these challenges in PolSAR feature extraction. Convolutional Neural Networks (CNNs) play a crucial role in capturing PolSAR image characteristics by leveraging kernel capabilities to consider local information and the complex-valued nature of PolSAR data. In this study, a novel three-branch fusion of complex-valued CNN named Shallow to Deep Feature Fusion Network (SDF2Net) is proposed for PolSAR image classification. To validate the performance of the proposed method, classification results are compared against multiple state-of-the-art approaches using the Airborne Synthetic Aperture Radar (AIRSAR) datasets of Flevoland, San Francisco, and ESAR Oberpfaffenhofen dataset. The results indicate that the proposed approach demonstrates improvements in OA, with 1.3% and 0.8% enhancement for the AIRSAR datasets and 0.5% improvement for the ESAR dataset. The analyses conducted on the Flevoland data underscore the effectiveness of the SDF2Net model, revealing a promising OA of 96.01% even with only 1% sampling ratio.

Recent state-of-the-art scene text recognition methods have primarily focused on horizontal text in images. However, in several Asian countries, including China, large amounts of text in signs, books, and TV commercials are vertically directed. Because the horizontal and vertical texts exhibit different characteristics, developing an algorithm that can simultaneously recognize both types of text in real environments is necessary. To address this problem, we adopted the direction encoding mask (DEM) and selective attention network (SAN) methods based on supervised learning. DEM contains directional information to compensate in cases that lack text direction; therefore, our network is trained using this information to handle the vertical text. The SAN method is designed to work individually for both types of text. To train the network to recognize both types of text and to evaluate the effectiveness of the designed model, we prepared a new synthetic vertical text dataset and collected an actual vertical text dataset (VTD142) from the Web. Using these datasets, we proved that our proposed model can accurately recognize both vertical and horizontal text and can achieve state-of-the-art results in experiments using benchmark datasets, including the street view test (SVT), IIIT-5k, and ICDAR. Although our model is relatively simple as compared to its predecessors, it maintains the accuracy and is trained in an end-to-end manner.

Acoustic energy mapping provides the functionality to obtain characteristics of acoustic sources, such as: presence, localization, type and trajectory of sound sources. Several beamforming-based techniques can be used for this purpose, however, they rely on the difference of arrival times of the signal at each capture node (or microphone), so it is of major importance to have synchronized multi-channel recordings. A Wireless Acoustic Sensor Network (WASN) can be very practical to install when used for mapping the acoustic energy of a given acoustic environment. However, they are known for having low synchronization between the recordings from each node. The objective of this paper is to characterize the impact of current popular synchronization methodologies as part of the WASN to capture reliable data to be used for acoustic energy mapping. The two evaluated synchronization protocols are: Network Time Protocol (NTP) y Precision Time Protocol (PTP). Additionally, three different audio capture methodologies were proposed for the WASN to capture the acoustic signal: two of them, recording the data locally and one sending the data through a local wireless network. As a real-life evaluation scenario, a WASN was built using nodes conformed by a Raspberry Pi 4B+ with a single MEMS microphone. Experimental results demonstrate that the most reliable methodology is using the PTP synchronization protocol and audio recording locally.

Purpose: This study aims to provide a foundational understanding of patient-centric perceptions of telemedicine implementation in rural Minnesota. Methods: This retrospective cohort study utilized a mail questionnaire to assess patient attitudes and behaviors regarding telemedicine during the emergency response to COVID-19. The target population was patients of a Primary Care clinic in eastern Minnesota. Descriptive statistics were used to assess respondents’ level of agreement with survey items scored on a 0-10 scale. Results: A majority of the respondents indicated that they were satisfied with using telemedicine (73.96%); that they gained an additional sense of control over their health condition due to the availability of telemedicine (52.08%); that they would support the use of telemedicine in the future (61.05%); that telemedicine increased access to care (65.63%); telemedicine was preferred over the clinic (61.46%), and that they would purchase new equipment to increase the utility of telemedicine services (54.17%). 51.04% of the respondents indicated a 10/10 level of satisfaction with telemedicine services. 48.96% indicated the strongest level of disagreement related to difficulty using telemedicine or a preference for in-person clinical appointments. Conclusions: Patients were found to have favorable perceptions of telemedicine overall. Though providers continue to recognize the benefit of face-to-face visits, their patients show an increased predilection for virtual care and telemedicine will likely remain an essential tool for providers to reach patients who would otherwise choose not to seek care.

The focus of this research is to analyze the results of encrypting audio using various authenticated encryption algorithms implemented in the Python cryptography library for ensuring authenticity and confidentiality of the original contents. The Advanced Encryption Standard (AES) is used as the underlying cryptographic primitive in conjunction with various modes including Galois Counter Mode (GCM), Counter with Cipher Block Chaining Message Authentication Code (CCM), and Cipher Block Chaining (CBC) with Keyed-Hashing for encrypting a relatively small audio file. The resulting encrypted audio shows similarity in the variance when encrypting using AES-GCM and AES-CCM. There is a noticeable reduction in variance of the performed encodings and an increase in the amount of time it takes to encrypt and decrypt the same audio file using AES-CBC with Keyed-Hashing. In addition, the corresponding encrypted using this mode audio spans a longer duration. As a result, AES should either have GCM or CCM for an efficient and reliable authenticated encryption integration within a workflow.

Audio recording in classrooms is a common practice in educational research, with applications ranging from detecting classroom activities to analysing student behaviour. Previous research has employed neural networks for classroom activity detection and speaker role identification. However, these recordings are often affected by background noise that can hinder further analysis, and the literature has only sought to identify noise with general filters and not specifically designed for classrooms. Although the use of high-end microphones and environmental monitoring can mitigate this problem, these solutions can be costly and potentially disruptive to the natural classroom environment. In this context, we propose the development of a neural network model that can specifically detect and filter out background noise in classroom recordings. This model would allow the use of lower quality recordings without compromising analysis capability, thus facilitating data collection in natural educational environments and reducing the costs associated with high-end recording equipment.

This study investigates anomaly detection of IoT device traffic using Convolutional Neural Networks (CNN) and Variational Autoencoders (VAE) to enhance the detection capability of security threats in IoT environments. A series of hardware configurations, software environments, and hyperparameters were utilized to optimize the training and testing processes of the models. The CNN model demonstrates robust classification performance, achieving an accuracy rate of 95.85% on the test dataset, effectively distinguishing between different types of IoT device traffic. Meanwhile, the VAE model exhibits proficient anomaly detection capabilities by effectively capturing abnormal patterns in the data using reconstruction loss and KL divergence. The combined use of CNN and VAE models offers a comprehensive solution to cybersecurity challenges in IoT environments. Future research directions include exploring diverse IoT traffic data, practical deployment for validation, and further optimization of model structures and parameters to improve performance and applicability.

This study employs the reflection symmetry decomposition (RSD) method to extract polarization scattering features from ground object images, aiming to determine the optimal data input scheme for deep learning networks in polarimetric synthetic aperture radar classification. Eight distinct polarizing feature combinations are designed, and the classification accuracy of various approaches is evaluated using the classic convolutional neural networks(CNNs) AlexNet and VGG16. The findings reveal that the commonly employed 6-parameter input scheme, favored by many researchers, lacks the comprehensive utilization of polarization information and warrants attention. Intriguingly, leveraging the complete 9-parameter input scheme based on the polarization coherence matrix results in improved classification accuracy. Furthermore, the input scheme incorporating all 21 parameters from the RSD and polarization coherence matrix notably enhances overall accuracy and the Kappa coefficient compared to the other 7 schemes. This comprehensive approach maximizes the utilization of polarization scattering information from ground objects, emerging as the most effective CNN input data scheme in this study. Additionally, the classification performance using the second and third component total power values (P2 and P3) from the RSD surpasses the approach utilizing surface scattering power value (PS) and secondary scattering power value (PD) from the same decomposition.

The dynamics of surface and sub-surface water events can lead to slope instability resulting in anomalies such as slough slides on earthen levees. Early detection of these anomalies by a remote sensing approach could save time versus direct assessment. We have implemented a supervised Mahalanobis distance classification algorithm for the detection of slough slides on levees using complex polarimetric Synthetic Aperture Radar (polSAR) data. The classifier output was followed by a spatial majority filter post-processing step which improved the accuracy. The effectiveness of the algorithm is demonstrated using fully quad-polarimetric L-band Synthetic Aperture Radar (SAR) imagery from the NASA Jet Propulsion Laboratory’s (JPL’s) Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR). The study area is a section of the lower Mississippi River valley in the southern USA. Slide detection accuracy of up to 98 percent was achieved, although the number of available slides examples was small.

Detecting and understanding emotions is critical for our daily activities. As emotion recognition (ER) systems develop, we start looking at more difficult cases than just acted adult audio-visual speech. In this work, we investigate automatic classification of audio-visual emotional speech of children. Our interest is, specifically, in the improvement of the utilization of the cross-modal relationships between the selected modalities: video and audio. To underscore the importance of developing ER systems for the real-world environment, we present a corpus of children’s emotional audio-visual speech that we collected. We select a state-of-the-art model as a baseline for the purposes of comparison and present several modifications focused on a deeper learning of the cross-modal relationships. By conducting experiments with our proposed approach and the selected baseline model, we observe a relative improvement in performance by 2%. Finally, we conclude that focusing more on the cross-modal relationships may be beneficial for building ER systems for child-machine communications and the environments where qualified professionals work with children.

For the Remotely Piloted Aircraft Systems (RPAS) market to continue its current growth rate, cost-effective "Detect and Avoid" systems that enable safe beyond visual line of sight (BVLOS) operations are critical. We propose an audio-based "Detect and Avoid" system, composed of microphones and an embedded computer, which performs real-time inferences using a sound event detection (SED) deep learning model. Two state-of-the-art SED models, YAMNet and VGGish, are fine-tuned using our dataset of aircraft sounds and their performances are compared for a wide range of configurations. YAMNet, whose MobileNet architecture is designed for embedded applications, outperformed VGGish both in terms of aircraft detection and computational performance. YAMNet's optimal configuration, with > 70% true positive rate and precision, results from combining data augmentation and undersampling with the highest available inference frequency (i.e. 10 Hz). While our proposed "Detect and Avoid" system already allows the detection of small aircraft from sound in real time, additional testing using multiple aircraft types is required. Finally, a larger training dataset, sensor fusion, or remote computations on cloud-based services could further improve system performance.

Recent advances in satellite technology have brought enormous potential to ecosystem mapping, which is one of the fundamental components of environmental studies. In this paper, a Random Forest classifier is applied for the strict assessment of the efficiency of ecosystem mapping through a detailed comparative analysis between combined Sentinel-1 and Sentinel-2 data and stand-alone Sentinel-2 imagery over three priority ecosystems, including wetlands, riverine areas, and mangroves in Bangladesh. The collocated images, based on the integration of Sentinel-1 data with Sentinel-2 data, would do better than Sentinel-2 imagery alone over various ecosystems. Particularly, in this study, attention focused on the Hakaluki Haor area for the wetlands, the Padma-Jamuna River confluence for the riverine ecosystem, and the Sundarban forest for mangroves. By leveraging Synthetic Aperture Radar (SAR) data in C-band dual-polarization from Sentinel-1 and four spectral bands (blue, green, red, and near-infrared) from Sentinel-2, the study analyzes imagery from December 2022 to February 2023. A 5% cloud masking filter is applied to optical data to enhance accuracy. In this methodology, 70% of the total signature values are used for training the classification model and the remaining 30% for testing. It can be noticed from the results that with the use of fused data, remarkably high accuracy in classification has been improved, such as overall accuracies of 94.17% for mangroves, 87.30% for riverine, and 85.96% for wetland ecosystems. In contrast, the use of singular Sentinel-2 imagery yields lower accuracies of 91.56%, 85.21%, and 82.51% for the respective ecosystems. The integration of radar data is shown to provide critical information, especially in environments with dense vegetation or cloud cover, where optical data alone may be insufficient. The findings of this study underline the limitations of relying on Sentinel-2 imagery to capture complex details of diverse ecosystems and highlight the need to include Sentinel-1 data for a more holistic analysis. This fusion allows improved accuracy to be achieved, which not only brings in more depth of ecological knowledge but also underpins more effective conservation strategies.

In recent times, BERT based transformer models have become an inseparable part of the 'tech stack' of text processing models. Similar progress is being observed in the speech domain with a multitude of models observing state-of-the-art results by using audio transformer models to encode speech. This begs the question of what are these audio transformer models learning. Moreover, although the standard methodology is to choose the last layer embedding for any downstream task, but is it the optimal choice? We try to answer these questions for the two recent audio transformer models, Mockingjay and wave2vec2.0. We compare them on a comprehensive set of language delivery and structure features including audio, fluency and pronunciation features. Additionally, we probe the audio models' understanding of textual surface, syntax, and semantic features and compare them to BERT. We do this over exhaustive settings for native, non-native, synthetic, read and spontaneous speech datasets

The growth rate in academic and scientific publications increased very fast. Researchers must choose a good representative and significant literature in publications, which has become challenging worldwide. Usually, the paper citation number indicates this paper's potential influence and importance. However, this standard metric of citation numbers cannot be used as a measurement to judge the popularity and significance of the recently published papers. To address this challenge, this paper presents an effective prediction method called FutureCite to predict the future citation level of research articles. FutureCite combines machine learning with text and graph mining techniques, leveraging their abilities in classification, datasets in-depth analysis, and feature extraction. FutureCite aims to predict future citation levels of research articles applying a multilabel classification approach. FutureCite can extract significant semantic features and capture the interconnection relationships found in scientific articles during feature extraction using textual content, citation networks, and metadata as feature resources. Our goal is to contribute to the advancement of substantial effective approaches impacting the citation counts in scientific publications by enhancing the precision of future citations. We conducted several experiments using a comprehensive publication dataset to evaluate our method and determine the impact of using a variety of machine learning algorithms. FutureCite demonstrated its robustness and efficiency and showed promising results based on different evaluation metrics. Using the FutureCite model has significant implications in improving the researchers' ability to determine targeted literature for their research and better understand the potential impact of research publications.

With the rise of piano teaching in recent years, many people have joined the team of piano learners. However, the expensive cost of manual instruction and the unique one-on-one teaching model have made piano learning an extravagant event. Most existing approaches based on the audio modality aim to evaluate piano players' skills. Unfortunately, these methods ignored the information contained in the video, which led to a one-sided and simplistic evaluation of the piano player's skills. More recently, multimodal-based methods are proposed to assess the skill level of piano players using both video and audio information. However, existing multimodal approaches use shallow networks to extract video and audio features, which are deficient in extracting complex spatio-temporal and time-frequency features from piano performance. Furthermore, the fingering and the pitch-rhythm information of the piano performance is contained in the spatio-temporal and time-frequency features, respectively. In the paper, we propose a ResNet-based audio-visual fusion model that combines video and audio features to assess the skill level of piano players. Firstly, ResNet18-3D is used as the backbone network for our visual branches, which can extract feature information from the video data. Then, we consider ResNet18-2D as the backbone network of the aural branch and extract the feature information from the audio data. The extracted video features are fused with the audio features to generate multimodal features for the final piano skill evaluation. The experimental results on the PISA dataset show that our proposed audio-visual fusion model, with a validation accuracy of 70.80%, outperforms the state-of-the-art methods in both performance and efficiency. Then, we also explore the impact of different layers of ResNet on model performance, and the experimental results show that the audio-visual fusion model dealing with the piano skill assessment problem can make full use of both feature information when the number of video features is close to the number of audio features.

This study aimed to assess the feasibility of employing audio and vision technologies as an integrated monitoring system to monitor the health and behavior of pigs under different environmental conditions and to associate these factors. A total of 81 growing pigs ((Landrace x Largewhite) x Duroc) with distinct weight categories (10.47 ± 2.57 kg and 28.81 ± 5.03 kg) were distributed into two distinct housing conditions: Control (characterized by low NH3 levels and low stocking density) and Treatment (marked by elevated NH3 levels and varying stocking densities). Each house was installed with a SoundTalks monitor for automated daily evaluation of respiratory health status (ReHS), and a camera with a microphone to facilitate manual tracking of respiratory symptoms and behavioral patterns. Results showed that the Treatment group encountered significantly high room temperature, NH3 concentration, and carbon dioxide levels, resulting in compromised growth performance—a phenomenon further exacerbated in high stocking density conditions. Pigs within the treatment exhibited increases (p < 0.05) in lateral and total lying behaviors, and incidences of ear and tail injuries, increases screaming frequency, conjunctivitis incidences, coughing frequency, elevated respiration rates, and a decrease in ReHS. The environmental conditions affect the behavior and health of the pigs. Moreover, the behavioral patterns of the pigs are associated with their health conditions. The SoundTalks system did not trigger any warnings during the experimental period. Nevertheless, trend analysis indicated a significant reduction in respiratory health in the treatment. In conclusion, this study underscores the efficacy of merging audio and visual technologies to holistically monitor pig health and behavior, enabling enhanced management strategies. Findings emphasize that monitoring the ReHS trend serves as a pivotal marker for identifying respiratory health problems, complementing the system's innate alarm functions.

Text-based CAPTCHA remains the most widely adopted security scheme, which is the first barrier to securing websites. Deep learning methods, especially Convolutional Neural Networks (CNNs) are the mainstream approach for text-CAPTCHA recognition, which are widely used in CAPTCHA vulnerability assessment and data collection. However, verification code recognizers are mostly deployed on the CPU platform as part of a web crawler and security assessment, they are required to have both low complexity and high recognition accuracy. Due to the specifically designed anti-attack mechanisms like noise, interference, geometric deformation, twisting, rotation, and character adhesion in text CAPTCHAs, some characters are difficult to efficiently identify with high accuracy in these complex CAPTCHA images. This paper proposed a recognition model named Adaptive-CAPTCHA with a CRNN module and trainable and configurable filtering networks, which effectively handle the interference and learn the correlation between characters in CAPTCHAs to enhance recognition accuracy. Experimental results on two datasets of different complexity show that compared with the baseline model Deep-CAPTCHA, the number of parameters of our proposed model is reduced by about 70% and the recognition accuracy is improved by more than 10 percentage points in the two datasets. In addition, the proposed model has a faster training convergence speed.

In dementia, particularly Alzheimer's Disease (AD), communication challenges are evident, especially in vocabulary and pragmatic aspects. Affected individuals often use vague, non-specific words, and their speech lacks informative nouns and verbs, leading to imprecise communication. However, aspects like sentence structure, phonology, and articulation are believed to remain intact until later stages, though this view is debated in the research community. The rise of Large Language Models (LLMs) has made significant strides in various domains, including sentiment analysis and question-answering. These advancements have been applied to dementia research, with studies using LLMs to analyze textual data. Some research incorporates pauses in text to enhance performance, while others utilize transfer learning techniques. However, limited datasets for dementia detection pose challenges in training LLMs. Our research presents a novel approach to measuring the impact of in-text encoding strategies by embedding special characters within the text to enhance model performance and incorporating sequences and summaries of their frequency. Our best model achieves 0.88 and 0.86 in f1-score and accuracy, respectively, whereas the baseline has 0.42 and 0.56 in f1-score and accuracy.

Large Language Models (LLMs) have proved significant proficiency when comes to code generation especially in Structured Query Language (SQL) for databases and recent successful Text-to-SQL method involves fine-tuning pre-trained LLMs for SQL generation tasks. Transforming natural language text into SQL queries, has been attempted to solve with various learning techniques including Few-shot learning[1], fine tuning. In this paper we propose Supervised fine-tuning (SFT) as a better alternative for learning technique for text-to-SQL generation task using Code-Llama that pushes state of art accuracy on spider test suite to 89.6% on dev set which represent first instance of surpassing the earlier best-in-class with 5.5% higher score and 86.8% of exact match accuracy on dev set. Furthermore, we demonstrate that properly prompted LLM along with SFT provides far fewer hallucinations and much more robust LLM that can be used as a general tool for any text-to-SQL generation use case.