Search Results (17)

In recent years, the integration of smartphone technology with novel sensing technologies, Artificial Intelligence (AI), and Deep Learning (DL) algorithms has revolutionized crop pest and disease surveillance. Efficient and accurate diagnosis is crucial to mitigate substantial economic losses in agriculture caused by diseases and pests. An innovative Apple^® and Android™ mobile application for citizen science has been developed, to enable real-time detection and identification of plant leaf diseases and pests, minimizing their impact on horticulture, viticulture, and olive cultivation. Leveraging DL algorithms, this application facilitates efficient data collection on crop pests and diseases, supporting crop yield protection and cost reduction in alignment with the Green Deal goal for 2030 by reducing pesticide use. The proposed citizen science tool involves all Farm to Fork stakeholders and farm citizens in minimizing damage to plant health by insect and fungal diseases. It utilizes comprehensive datasets, including images of various diseases and insects, within a robust Decision Support System (DSS) where DL models operate. The DSS connects directly with users, allowing them to upload crop pest data via the mobile application, providing data-driven support and information. The application stands out for its scalability and interoperability, enabling the continuous integration of new data to enhance its capabilities. It supports AI-based imaging analysis of quarantine pests, invasive alien species, and emerging and native pests, thereby aiding post-border surveillance programs. The mobile application, developed using a Python-based REST API, PostgreSQL, and Keycloak, has been field-tested, demonstrating its effectiveness in real-world agriculture scenarios, such as detecting Tuta absoluta (Meyrick) infestation in tomato cultivations. The outcomes of this study in T. absoluta detection serve as a showcase scenario for the proposed citizen science tool’s applicability and usability, demonstrating a 70.2% accuracy (mAP₅₀) utilizing advanced DL models. Notably, during field testing, the model achieved detection confidence levels of up to 87%, enhancing pest management practices. Full article

The colors of walnut fruits and leaves are similar in the oil transformation period, and the fruits are easily blocked by the branches and leaves. On the basis of the improved YOLOv7-tiny, a detection model is proposed and integrated into an Android application to solve the problem of walnut identification. Ablation experiments conducted with three improved strategies show that the strategies can effectively enhance the performance of the model. In terms of combinatorial optimization, the YOLOv7-tiny detection model that combines FasterNet and LightMLP modules works excellently. Its AP₅₀ and AP_50–95 are 3.1 and 4 percentage points (97.4% and 77.3%, respectively) higher than those of the original model. YOLOv7-tiny’s model size and number of parameters are reduced by 14.6% and 14.4%, respectively, relative to those of the original model, and its detection time decreases to 15.4 ms. The model has good robustness and generalization ability and can provide a technical reference for intelligent real-time detection of walnuts during the oil conversion period. Full article

This review addresses the complexities of type 1 diabetes (T1D) and its associated complications, with a particular focus on diabetic retinopathy (DR). This review outlines the progression from non-proliferative to proliferative diabetic retinopathy and diabetic macular edema, highlighting the role of dysglycemia in the pathogenesis of these conditions. A significant portion of this review is devoted to technological advances in diabetes management, particularly the use of hybrid closed-loop systems (HCLSs) and to the potential of open-source HCLSs, which could be easily adapted to different patients’ needs using big data analytics and machine learning. Personalized HCLS algorithms that integrate factors such as patient lifestyle, dietary habits, and hormonal variations are highlighted as critical to reducing the incidence of diabetes-related complications and improving patient outcomes. Full article

Pig counting is an important work in the breeding process of large-scale pig farms. In order to achieve high-precision pig identification in the conditions of pigs occluding each other, illumination difference, multiscenes, and differences in the number of pigs and the imaging size, and to also reduce the number of parameters of the model, a pig counting algorithm of improved YOLOv5n was proposed. Firstly, a multiscene dataset is created by selecting images from several different pig farms to enhance the generalization performance of the model; secondly, the Backbone of YOLOv5n was replaced by the FasterNet model to reduce the number of parameters and calculations to lay the foundation for the model to be applied to Android system; thirdly, the Neck of YOLOv5n was optimized by using the E-GFPN structure to enhance the feature fusion capability of the model; Finally, Focal EIoU loss function was used to replace the CIoU loss function of YOLOv5n to improve the model’s identification accuracy. The results showed that the AP of the improved model was 97.72%, the number of parameters, the amount of calculation, and the size of the model were reduced by 50.57%, 32.20%, and 47.21% compared with YOLOv5n, and the detection speed reached 75.87 f/s. The improved algorithm has better accuracy and robustness in multiscene and complex pig house environments, which not only ensured the accuracy of the model but also reduced the number of parameters as much as possible. Meanwhile, a pig counting application for the Android system was developed based on the optimized model, which truly realized the practical application of the technology. The improved algorithm and application could be easily extended and applied to the field of livestock and poultry counting, such as cattle, sheep, geese, etc., which has a widely practical value. Full article

This research proposes a study on two-way communication between deaf/mute and normal people using an Android application. Despite advancements in technology, there is still a lack of mobile applications that facilitate two-way communication between deaf/mute and normal people, especially by using Bahasa Isyarat Malaysia (BIM). This project consists of three parts: First, we use BIM letters, which enables the recognition of BIM letters and BIM combined letters to form a word. In this part, a MobileNet pre-trained model is implemented to train the model with a total of 87,000 images for 29 classes, with a 10% test size and a 90% training size. The second part is BIM word hand gestures, which consists of five classes that are trained with the SSD-MobileNet-V2 FPNLite 320 × 320 pre-trained model with a speed of 22 s/frame rate and COCO mAP of 22.2, with a total of 500 images for all five classes and first-time training set to 2000 steps, while the second- and third-time training are set to 2500 steps. The third part is Android application development using Android Studio, which contains the features of the BIM letters and BIM word hand gestures, with the trained models converted into TensorFlow Lite. This feature also includes the conversion of speech to text, whereby this feature allows converting speech to text through the Android application. Thus, BIM letters obtain 99.75% accuracy after training the models, while BIM word hand gestures obtain 61.60% accuracy. The suggested system is validated as a result of these simulations and tests. Full article

Deep learning algorithms, such as convolutional neural networks (CNNs), have been widely studied and applied in various fields including agriculture. Agriculture is the most important source of food and income in human life. In most countries, the backbone of the economy is based on agriculture. Pests are one of the major challenges in crop production worldwide. To reduce the overall production and economic loss from pests, advancement in computer vision and artificial intelligence may lead to early and small pest detection with greater accuracy and speed. In this paper, an approach for early pest detection using deep learning and convolutional neural networks has been presented. Object detection is applied on a dataset with images of thistle caterpillars, red beetles, and citrus psylla. The input dataset contains 9875 images of all the pests under different illumination conditions. State-of-the-art Yolo v3, Yolov3-Tiny, Yolov4, Yolov4-Tiny, Yolov6, and Yolov8 have been adopted in this study for detection. All of these models were selected based on their performance in object detection. The images were annotated in the Yolo format. Yolov8 achieved the highest mAP of 84.7% with an average loss of 0.7939, which is better than the results reported in other works when compared to small pest detection. The Yolov8 model was further integrated in an Android application for real time pest detection. This paper contributes the implementation of novel deep learning models, analytical methodology, and a workflow to detect pests in crops for effective pest management. Full article

The agriculture sector plays a crucial role in supplying nutritious and high-quality food. Plant disorders significantly impact crop productivity, resulting in an annual loss of 33%. The early and accurate detection of plant disorders is a difficult task for farmers and requires specialized knowledge, significant effort, and labor. In this context, smart devices and advanced artificial intelligence techniques have significant potential to pave the way toward sustainable and smart agriculture. This paper presents a deep learning-based android system that can diagnose ginger plant disorders such as soft rot disease, pest patterns, and nutritional deficiencies. To achieve this, state-of-the-art deep learning models were trained on a real dataset of 4,394 ginger leaf images with diverse backgrounds. The trained models were then integrated into an Android-based mobile application that takes ginger leaf images as input and performs the real-time detection of crop disorders. The proposed system shows promising results in terms of accuracy, precision, recall, confusion matrices, computational cost, Matthews correlation coefficient (MCC), mAP, and F1-score. Full article

Received signal strength (RSS) has been one of the most used observables for location purposes due to its availability at almost every wireless device. However, the volatile nature of RSS tends to yield to non-reliable location solutions. IEEE 802.11mc enabled the use of the round trip time (RTT) for positioning, which is expected to be a more consistent observable for location purposes. This approach has been gaining support from several companies such as Google, which introduced that feature in the Android O.S. As a result, RTT estimation is now available in several recent off-the-shelf devices, opening a wide range of new approaches for computing location. However, RTT has been traditionally addressed to multilateration solutions. Few works exist that assess the feasibility of the RTT as an accurate feature in positioning methods based on classification algorithms. An attempt is made in this paper to fill this gap by investigating the performance of several classification models in terms of accuracy and positioning errors. The performance is assessed using different AP layouts, distinct AP vendors, and different frequency bands. The accuracy and precision of the RTT-based position estimation is always better than the one obtained with RSS in all the studied scenarios, and especially when few APs are available. In addition, all the considered ML algorithms perform pretty well. As a result, it is not necessary to use more complex solutions (e.g., SVM) when simpler ones (e.g., nearest neighbor classifiers) achieve similar results both in terms of accuracy and location error. Full article

Skin image analysis using artificial intelligence (AI) has recently attracted significant research interest, particularly for analyzing skin images captured by mobile devices. Acne is one of the most common skin conditions with profound effects in severe cases. In this study, we developed an AI system called AcneDet for automatic acne object detection and acne severity grading using facial images captured by smartphones. AcneDet includes two models for two tasks: (1) a Faster R-CNN-based deep learning model for the detection of acne lesion objects of four types, including blackheads/whiteheads, papules/pustules, nodules/cysts, and acne scars; and (2) a LightGBM machine learning model for grading acne severity using the Investigator’s Global Assessment (IGA) scale. The output of the Faster R-CNN model, i.e., the counts of each acne type, were used as input for the LightGBM model for acne severity grading. A dataset consisting of 1572 labeled facial images captured by both iOS and Android smartphones was used for training. The results show that the Faster R-CNN model achieves a mAP of 0.54 for acne object detection. The mean accuracy of acne severity grading by the LightGBM model is 0.85. With this study, we hope to contribute to the development of artificial intelligent systems to help acne patients better understand their conditions and support doctors in acne diagnosis. Full article

Indoor localization using fine time measurement (FTM) round-trip time (RTT) with respect to cooperating Wi-Fi access points (APs) has been shown to work well and provide 1–2 m accuracy in both 2D and 3D applications. This approach depends on APs implementing the IEEE 802.11-2016 (also known as IEEE 802.11mc) Wi-Fi standard (“two-sided” RTT). Unfortunately, the penetration of this Wi-Fi protocol has been slower than anticipated, perhaps because APs tend not to be upgraded as often as other kinds of electronics, in particular in large institutions—where they would be most useful. Recently, Google released Android 12, which also supports an alternative “one-sided” RTT method that will work with legacy APs as well. This method cannot subtract out the “turn-around” time of the signal, and so, produces distance estimates that have much larger offsets than those seen with two-sided RTT—and the results are somewhat less accurate. At the same time, this method makes possible distance measurements for many APs that previously could not be used. This increased accessibility can compensate for the decreased accuracy of individual measurements. We demonstrate here indoor localization using one-sided RTT with respect to legacy APs that do not support IEEE 802.11-2016. The accuracy achieved is 3–4 m in cluttered environments with few line-of-sight readings (and using only 20 MHz bandwidths). This is not as good as for two-sided RTT, where 1–2 m accuracy has been achieved (using 80 MHz bandwidths), but adequate for many applications A wider Wi-Fi channel bandwidth would increase the accuracy further. As before, Bayesian grid update is the preferred method for determining position and positional accuracy, but the observation model now is different from that for two-sided RTT. As with two-sided RTT, the probability of an RTT measurement below the true distance is very low, but, in the other direction, the range of measurements for a given distance can be much wider (up to well over twice the actual distance). We describe methods for formulating useful observation models. As with two-sided RTT, the offset or bias in distance measurements has to be subtracted from the reported measurements. One difference is that here, the offsets are large (typically in the 2400–2700 m range) because of the “turn-around time” of roughly 16 μs (i.e., about two orders of magnitude larger than the time of flight one is attempting to measure). We describe methods for estimating these offsets and for minimizing the effort required to do so when setting up an installation with many APs. Full article

The use of deep neural networks (DNNs) in plant phenotyping has recently received considerable attention. By using DNNs, valuable insights into plant traits can be readily achieved. While these networks have made considerable advances in plant phenotyping, the results are processed too slowly to allow for real-time decision-making. Therefore, being able to perform plant phenotyping computations in real-time has become a critical part of precision agriculture and agricultural informatics. In this work, we utilize state-of-the-art object detection networks to accurately detect, count, and localize plant leaves in real-time. Our work includes the creation of an annotated dataset of Arabidopsis plants captured using Cannon Rebel XS camera. These images and annotations have been complied and made publicly available. This dataset is then fed into a Tiny-YOLOv3 network for training. The Tiny-YOLOv3 network is then able to converge and accurately perform real-time localization and counting of the leaves. We also create a simple robotics platform based on an Android phone and iRobot create2 to demonstrate the real-time capabilities of the network in the greenhouse. Additionally, a performance comparison is conducted between Tiny-YOLOv3 and Faster R-CNN. Unlike Tiny-YOLOv3, which is a single network that does localization and identification in a single pass, the Faster R-CNN network requires two steps to do localization and identification. While with Tiny-YOLOv3, inference time, F1 Score, and false positive rate (FPR) are improved compared to Faster R-CNN, other measures such as difference in count (DiC) and AP are worsened. Specifically, for our implementation of Tiny-YOLOv3, the inference time is under 0.01 s, the F1 Score is over 0.94, and the FPR is around 24%. Last, transfer learning using Tiny-YOLOv3 to detect larger leaves on a model trained only on smaller leaves is implemented. The main contributions of the paper are in creating dataset (shared with the research community), as well as the trained Tiny-YOLOv3 network for leaf localization and counting. Full article

As Bring Your Own Device (BYOD) policy has become widely accepted in the enterprise, anyone with a mobile device that supports Wi-Fi tethering can provide an active wireless Internet connection to other devices without restriction from network administrators. Despite the potential benefits of Wi-Fi tethering, it raises new security issues. The open source nature of mobile operating systems (e.g., Google Android or OpenWrt) can be easily manipulated by selfish users to provide an unfair advantage throughput performance to their tethered devices. The unauthorized tethering can interfere with nearby well-planned access points (APs) within Wi-Fi networks, which results in serious performance problems. In this paper, we first conduct an extensive evaluation study and demonstrate that the abuse of Wi-Fi tethering that adjusts the clear channel access parameters has strong adverse effects in Wi-Fi networks, while providing the manipulated device a high throughput gain. Subsequently, an online detection scheme diagnoses the network condition and detects selfish tethering devices by passively exploiting the packet loss information of on-going transmissions. Our evaluation results show that the proposed method accurately distinguishes the manipulated tethering behavior from other types of misbehavior, including the hidden node problem. Full article

There are widely used standard clinical tests to estimate the instability of an anterior cruciate ligament (ACL) deficient knee by assessing the translation of the tibia with respect to the femur. However, the assessment of those tests could be quite subjective. The goal of this study is to present a universally affordable open-source Android application that is easy and quick. Moreover, it provides the possibility for a quantitative and objective analysis of that instability. The anterior–posterior knee translation of seven subjects was assessed using the open-source Android application developed. A single Android smartphone and the placement of three green skin adhesives are all that is required to use it. The application was developed using the image-processing features of the open-source OpenCV Library. An open-source Android application was developed to measure anterior–posterior (AP) translation in ACL-deficient subjects. The application identified differences in the AP translation between the ipsilateral and the contralateral legs of seven ACL-deficient subjects during Lachman and Pivot–Shift tests. Three out of seven subjects were under anesthesia. Those three were also the ones with significant differences. The application detected differences in the AP translation between the ipsilateral and contralateral legs of subjects with ACL deficiency. The use of the application represents an easy, low-cost, reliable and quick way to assess knee instability quantitatively. Full article

The fingerprint method has been widely adopted in Wi-Fi indoor positioning because of its advantage in non-line-of-sight channels between access points (APs) and mobile users. However, the received signal strength (RSS) during the fingerprint positioning process generally varies due to the dissimilar hardware configurations of heterogeneous smartphones. This difference may degrade the accuracy of fingerprint matching between fingerprint and test data. Thus, this paper puts forward a fingerprint method based on grey relational analysis (GRA) to approach the challenge of heterogeneous smartphones and to improve positioning accuracy. Initially, the grey relational coefficient (GRC) between the RSS comparability sequence of each reference point (RP) and the RSS reference sequence of the test point (TP) is calculated. Subsequently, the grey relational degree (GRD) between each RP and TP is determined on the basis of GRC, and the K most relational RPs are selected in accordance with the value of GRD. Finally, the user location is determined by weighting the K most relational RPs that correspond to the coordinates. The main advantage of this GRA method is that it does not require device calibration when handling heterogeneous smartphone problems. We further carry out extensive experiments using heterogeneous Android smartphones in an office environment to verify the positioning performance of the proposed method. Experimental results indicate that the proposed method outperforms the existing ones no matter whether heterogeneous smartphones are used. Full article

Smartphone indoor localization has attracted considerable attention over the past decade because of the considerable business potential in terms of indoor navigation and location-based services. In particular, Wi-Fi RSS (received signal strength) fingerprinting for indoor localization has received significant attention in the industry, for its advantage of freely using off-the-shelf APs (access points). However, RSS measured by heterogeneous mobile devices is generally biased due to the variety of embedded hardware, leading to a systematical mismatch between online measures and the pre-established radio maps. Additionally, the fingerprinting method based on a single RSS measurement usually suffers from signal fluctuations due to environmental changes or human body blockage, leading to possible large localization errors. In this context, this study proposes a space-constrained pairwise signal strength differences (PSSD) strategy to improve Wi-Fi fingerprinting reliability, and mitigate the effect of hardware bias of different smartphone devices on positioning accuracy without requiring a calibration process. With the efforts of these two aspects, the proposed solution enhances the usability of Wi-Fi fingerprint positioning. The PSSD approach consists of two critical operations in constructing particular fingerprints. First, we construct the signal strength difference (SSD) radio map of the area of interest, which uses the RSS differences between APs to minimize the device-dependent effect. Then, the pairwise RSS fingerprints are constructed by leveraging the time-series RSS measurements and potential spatial topology of pedestrian locations of these measurement epochs, and consequently reducing possible large positioning errors. To verify the proposed PSSD method, we carry out extensive experiments with various Android smartphones in a campus building. In the case of heterogeneous devices, the experimental results demonstrate that PSSD fingerprinting achieves a mean error ∼20% less than conventional RSS fingerprinting. In addition, PSSD fingerprinting achieves a 90-percentile accuracy of no greater than 5.5 m across the tested heterogeneous smartphones Full article