Search Results (306)

Background: Smartphones offer a promising tool for monitoring refractive error, especially in underserved areas where there is a shortage of eye-care professionals. We propose a novel method for measuring spherical over-refraction using smartphones. Methods: Specific levels of myopia using positive spherical trial lenses, ranging from 0.00 D to 1.50 D in 0.25 D increments, were induced in 30 young participants (22 ± 5 years). A comparison was conducted between the induced over-refraction and the measurements obtained using a non-commercial mobile application based on the face–device distance measurement using the front camera while the subject was performing a resolution task. Results: Calibrated mobile app over-refraction results showed that 89.5% of the estimates had an error ≤ 0.25 D, and no errors exceeding 0.50 D. Bland–Altman analysis revealed no significant bias between app and clinical over-refraction, with a mean difference of 0.00 D ± 0.44 D (p = 0.981), indicating high accuracy and precision of the method. Conclusions: The methodology used shows high accuracy and precision in the measurement of the spherical over-refraction with only the use of a smartphone, allowing self-monitorization of potential myopia progression. Full article

With the rapid proliferation of artificial intelligence (AI) applications, an increasing number of edge devices—such as smartphones, cameras, and embedded controllers—are being tasked with performing AI-based inference. Due to constraints in storage capacity, computational power, and network connectivity, these devices are often categorized as operating in resource-constrained environments. In such scenarios, deploying powerful Transformer-based models like ChatGPT and Vision Transformers is highly impractical because of their large parameter sizes and intensive computational requirements. While lightweight Transformer models, such as MobileViT, offer a promising solution to meet storage and computational limitations, their robustness remains insufficient. This poses a significant security risk for AI applications, particularly in critical edge environments. To address this challenge, our research focuses on enhancing the robustness of lightweight Transformer models under resource-constrained conditions. First, we propose a comprehensive robustness evaluation framework tailored for lightweight Transformer inference. This framework assesses model robustness across three key dimensions: noise robustness, distributional robustness, and adversarial robustness. It further investigates how model size and hardware limitations affect robustness, thereby providing valuable insights for robustness-aware model design. Second, we introduce a novel adversarial robustness enhancement strategy that integrates lightweight modeling techniques. This approach leverages methods such as gradient clipping and layer-wise unfreezing, as well as decision boundary optimization techniques like TRADES and SMART. Together, these strategies effectively address challenges related to training instability and decision boundary smoothness, significantly improving model robustness. Finally, we deploy the robust lightweight Transformer models in real-world resource-constrained environments and empirically validate their inference robustness. The results confirm the effectiveness of our proposed methods in enhancing the robustness and reliability of lightweight Transformers for edge AI applications. Full article

(1) Background: A severe decline in knee joint function significantly affects the mobility of the elderly, making it a key concern in the field of geriatric health. To alleviate the pressure on the knee joints of the elderly during daily movements such as sitting and standing, effective biomechanical solutions are required. (2) Methods: In this study, a biomechanical framework was established based on mechanical analysis to derive the transfer relationship between the ground reaction force and the knee joint moment. Experiments were designed to collect knee joint data on the elderly during the sit-to-stand process. Meanwhile, magnetic resonance imaging (MRI) images were processed through a medical imaging control system to construct a detailed digital 3D knee joint model. A finite element analysis was used to verify the model to ensure the accuracy of its structure and mechanical properties. An improved radial basis function was used to fit the pressure during the entire sit-to-stand conversion process to reduce the computational workload, with an error of less than 5%. In addition, a small-target human key point recognition network was developed to analyze the image sequences captured by the camera. The knee joint angle and the knee joint pressure distribution during the sit-to-stand conversion process were mapped to a three-dimensional interactive platform to form a digital twin system. (3) Results: The system can effectively capture the biomechanical behavior of the knee joint during movement and shows high accuracy in joint angle tracking and structure simulation. (4) Conclusions: This study provides an accurate and comprehensive method for analyzing the biomechanical characteristics of the knee joint during the movement of the elderly, laying a solid foundation for clinical rehabilitation research and the design of assistive devices in the field of rehabilitation medicine. Full article

The present research is related to the application of artificial intelligence (AI) approaches for classifying surface textures, specifically regular reliefs patterns formed by ball burnishing operations. A two-stage methodology is employed, starting with the creation of regular reliefs (RRs) on test parts by ball burnishing, followed by 3D topography scanning with Alicona device and data preprocessing with Gwyddion, and Blender software, where the acquired 3D topographies are converted into a set of 2D images, using various virtual camera movements and lighting to simulate the symmetrical fluctuations around the tool-path of the real camera. Four pre-trained convolutional neural networks (DenseNet121, EfficientNetB0, MobileNetV2, and VGG16) are used as a base for transfer learning and tested for their generalization performance on different combinations of synthetic and real image datasets. The models were evaluated by using confusion matrices and four additional metrics. The results show that the pretrained VGG16 model generalizes the best regular reliefs textures (96%), in comparison with the other models, if it is subjected to transfer learning via feature extraction, using mixed dataset, which consist of 34,037 images in following proportions: non-textured synthetic (87%), textured synthetic (8%), and real captured (5%) images of such a regular relief. Full article

The growing implementation of digital platforms and mobile devices in educational environments has generated the need to explore new approaches for evaluating the learning experience beyond traditional self-reports or instructor presence. In this context, the NPFC-Test dataset was created from an experimental protocol conducted at the Experiential Classroom of the Institute for the Future of Education. The dataset was built by collecting multimodal indicators such as neuronal, physiological, and facial data using a portable EEG headband, a medical-grade biometric bracelet, a high-resolution depth camera, and self-report questionnaires. The participants were exposed to a digital test lasting 20 min, composed of audiovisual stimuli and cognitive challenges, during which synchronized data from all devices were gathered. The dataset includes timestamped records related to emotional valence, arousal, and concentration, offering a valuable resource for multimodal learning analytics (MMLA). The recorded data were processed through calibration procedures, temporal alignment techniques, and emotion recognition models. It is expected that the NPFC-Test dataset will support future studies in human–computer interaction and educational data science by providing structured evidence to analyze cognitive and emotional states in learning processes. In addition, it offers a replicable framework for capturing synchronized biometric and behavioral data in controlled academic settings. Full article

This paper presents a novel approach for predicting Particulate Matter (PM) concentrations using mobile camera devices. In response to persistent air pollution challenges across Japan, we developed a system that utilizes cutting-edge transformer-based deep learning architectures to estimate PM values from imagery captured by smartphone cameras. Our approach employs Contrastive Language–Image Pre-Training (CLIP) as a multimodal framework to extract visual features associated with PM concentration from environmental scenes. We first developed a baseline through comparative analysis of time-series models for 1D PM signal prediction, finding that linear models, particularly NLinear, outperformed complex transformer architectures for short-term forecasting tasks. Building on these insights, we implemented a CLIP-based system for 2D image analysis that achieved a Top-1 accuracy of 0.24 and a Top-5 accuracy of 0.52 when tested on diverse smartphone-captured images. The performance evaluations on Graphics Processing Unit (GPU) and Single-Board Computer (SBC) platforms highlight a viable path toward edge deployment. Processing times of 0.29 s per image on the GPU versus 2.68 s on the SBC demonstrate the potential for scalable, real-time environmental monitoring. We consider that this research connects high-performance computing with energy-efficient hardware solutions, creating a practical framework for distributed environmental monitoring that reduces reliance on costly centralized monitoring systems. Our findings indicate that transformer-based multimodal models present a promising approach for mobile sensing applications, with opportunities for further improvement through seasonal data expansion and architectural refinements. Full article

Small object detection in traffic scenes presents unique challenges for mobile robots operating under constrained computational resources and highly dynamic environments. Unlike general object detection, small targets often suffer from low resolution, weak semantic cues, and frequent occlusion, especially in complex outdoor scenarios. This study systematically analyses the challenges, technical advances, and deployment strategies for small object detection tailored to mobile robotic platforms. We categorise existing approaches into three main strategies: feature enhancement (e.g., multi-scale fusion, attention mechanisms), network architecture optimisation (e.g., lightweight backbones, anchor-free heads), and data-driven techniques (e.g., augmentation, simulation, transfer learning). Furthermore, we examine deployment techniques on embedded devices such as Jetson Nano and Raspberry Pi, and we highlight multi-modal sensor fusion using Light Detection and Ranging (LiDAR), cameras, and Inertial Measurement Units (IMUs) for enhanced environmental perception. A comparative study of public datasets and evaluation metrics is provided to identify current limitations in real-world benchmarking. Finally, we discuss future directions, including robust detection under extreme conditions and human-in-the-loop incremental learning frameworks. This research aims to offer a comprehensive technical reference for researchers and practitioners developing small object detection systems for real-world robotic applications. Full article

Facial authentication has gained widespread adoption as a biometric authentication method, offering a convenient alternative to traditional password-based systems, particularly on mobile devices equipped with front-facing cameras. While this technology enhances usability and security by eliminating password management, it remains highly susceptible to spoofing attacks. Adversaries can exploit facial recognition systems using pre-recorded photos, videos, or even sophisticated 3D models of victims’ faces to bypass authentication mechanisms. The increasing availability of personal images on social media further amplifies this risk, making robust anti-spoofing mechanisms essential for secure facial authentication. To address these challenges, we introduce FaceCloseup, a novel liveness detection technique that strengthens facial authentication by leveraging perspective distortion inherent in close-up shots of real, 3D faces. Instead of relying on additional sensors or user-interactive gestures, FaceCloseup passively analyzes facial distortions in video frames captured by a mobile device’s camera, improving security without compromising user experience. FaceCloseup effectively distinguishes live faces from spoofed attacks by identifying perspective-based distortions across different facial regions. The system achieves a 99.48% accuracy in detecting common spoofing methods—including photo, video, and 3D model-based attacks—and demonstrates 98.44% accuracy in differentiating between individual users. By operating entirely on-device, FaceCloseup eliminates the need for cloud-based processing, reducing privacy concerns and potential latency in authentication. Its reliance on natural device movement ensures a seamless authentication experience while maintaining robust security. Full article

This study presents a novel investigation into immersive teleoperation systems using collaborative, device-agnostic interfaces for advancing smart haptics in industrial assistive applications. The research focuses on evaluating the quality of experience (QoE) of users interacting with a teleoperation system comprising a local robotic arm, a robot gripper, and heterogeneous remote tracking and haptic feedback devices. By employing a modular device-agnostic framework, the system supports flexible configurations, including one-user-one-equipment (1U-1E), one-user-multiple-equipment (1U-ME), and multiple-users-multiple-equipment (MU-ME) scenarios. The experimental set-up involves participants manipulating predefined objects and placing them into designated baskets by following specified 3D trajectories. Performance is measured using objective QoE metrics, including temporal efficiency (time required to complete the task) and spatial accuracy (trajectory similarity to the predefined path). In addition, subjective QoE metrics are assessed through detailed surveys, capturing user perceptions of presence, engagement, control, sensory integration, and cognitive load. To ensure flexibility and scalability, the system integrates various haptic configurations, including (1) a Touch kinaesthetic device for precision tracking and grounded haptic feedback, (2) a DualSense tactile joystick as both a tracker and mobile haptic device, (3) a bHaptics DK2 vibrotactile glove with a camera tracker, and (4) a SenseGlove Nova force-feedback glove with VIVE trackers. The modular approach enables comparative analysis of how different device configurations influence user performance and experience. The results indicate that the objective QoE metrics varied significantly across device configurations, with the Touch and SenseGlove Nova set-ups providing the highest trajectory similarity and temporal efficiency. Subjective assessments revealed a strong correlation between presence and sensory integration, with users reporting higher engagement and control in scenarios utilizing force feedback mechanisms. Cognitive load varied across the set-ups, with more complex configurations (e.g., 1U-ME) requiring longer adaptation periods. This study contributes to the field by demonstrating the feasibility of a device-agnostic teleoperation framework for immersive industrial applications. It underscores the critical interplay between objective task performance and subjective user experience, providing actionable insights into the design of next-generation teleoperation systems. Full article

Cataract remains the leading cause of blindness worldwide, and the number of individuals affected by this condition is expected to rise significantly due to global population ageing. Early diagnosis is crucial, as delayed treatment may result in irreversible vision loss. This study explores and presents a mobile application for Android devices designed for the detection of cataracts using deep learning models. The proposed solution utilizes a multi-stage classification approach to analyze ocular images acquired with a slit lamp, sourced from the Nuclear Cataract Database for Biomedical and Machine Learning Applications. The process involves identifying pathological features and assessing the severity of the detected condition, enabling comprehensive characterization of the NC (nuclear cataract) of cataract progression based on the LOCS III scale classification. The evaluation included a range of convolutional neural network architectures, from larger models like VGG16 and ResNet50, to lighter alternatives such as VGG11, ResNet18, MobileNetV2, and EfficientNet-B0. All models demonstrated comparable performance, with classification accuracies exceeding 91–94.5%. The trained models were optimized for mobile deployment, enabling real-time analysis of eye images captured with the device camera or selected from local storage. The presented mobile application, trained and validated on authentic clinician-labeled pictures, represents a significant advancement over existing mobile tools. The preliminary evaluations demonstrated a high accuracy in cataract detection and severity grading. These results confirm the approach is feasible and will serve as the foundation for ongoing development and extensions. Full article

The increasing volume of global waste presents a critical environmental and societal challenge, demanding innovative solutions to support sustainable practices such as recycling. Advances in Computer Vision (CV) have enabled automated waste recognition systems that guide users in correctly sorting their waste, with state-of-the-art architectures achieving high accuracy. More recently, attention has shifted toward lightweight and efficient models suitable for mobile and edge deployment. These systems process data from integrated camera sensors in Internet of Things (IoT) devices, operating in real time to classify waste at the point of disposal, whether embedded in smart bins, mobile applications, or assistive tools for household use. In this work, we extend our previous research by improving both dataset diversity and model efficiency. We introduce an expanded dataset that includes an organic waste class and more heterogeneous images, and evaluate a range of quantized CNN models to reduce inference time and resource usage. Additionally, we explore ensemble strategies using aggregation functions to boost classification performance, and validate selected models on real embedded hardware and under simulated lighting variations. Our results support the development of robust, real-time recycling assistants for resource-constrained devices. We also propose architectural deployment scenarios for smart containers, and cloud-assisted solutions. By improving waste sorting accuracy, these systems can help reduce landfill use, support citizen engagement through real-time feedback, increase material recovery, support data-informed environmental decision making, and ease operational challenges for recycling facilities caused by misclassified materials. Ultimately, this contributes to circular economy objectives and advances the broader field of environmental intelligence. Full article

In this study, we demonstrate an application for 5G networks in mobile and remote GPR scanning situations to detect buried objects by experts while the operator is performing the scans. Using a GSSI SIR-30 system in conjunction with the RealSense camera for visual mapping of the surveyed area, subsurface GPR scans were created and transmitted for remote processing. Using mobile networks, the raw B-scan files were transmitted at a sufficient rate, a maximum of 0.034 ms mean latency, to enable near real-time edge processing. The performance of 5G networks in handling the data transmission for the GPR scans and edge computing was compared to the performance of 4G networks. In addition, long-range low-power devices, namely Wi-Fi HaLow and Wi-Fi hotspots, were compared as local alternatives to cellular networks. Augmented reality headset representation of the F-scans is proposed as a method of assisting the operator in using the edge-processed scans. These promising results bode well for the potential of remote processing of GPR data in augmented reality applications. Full article

Communication barriers between deaf and hearing individuals present significant challenges to social inclusion, highlighting the need for effective technological aids. This study aimed to bridge this gap by developing a mobile system for the real-time translation of Portuguese Sign Language (LGP) alphabet gestures into text, addressing a specific technological void for LGP. The core of the solution is a mobile application integrating two distinct machine learning approaches trained on a custom LGP dataset: firstly, a Convolutional Neural Network (CNN) optimized with TensorFlow Lite for efficient, on-device image classification, enabling offline use; secondly, a method utilizing MediaPipe for hand landmark extraction from the camera feed, with classification performed by a server-side Multilayer Perceptron (MLP). Evaluation tests confirmed that both approaches could recognize LGP alphabet gestures with good accuracy (F1-scores of approximately 76% for the CNN and 77% for the MediaPipe+MLP) and processing speed (1 to 2 s per gesture on high-end devices using the CNN and 3 to 5 s under typical network conditions using MediaPipe+MLP), facilitating efficient real-time translation, though performance trade-offs regarding speed versus accuracy under different conditions were observed. The implementation of this dual-method system provides crucial flexibility, adapting to varying network conditions and device capabilities, and offers a scalable foundation for future expansion to include more complex gestures. This work delivers a practical tool that may contribute to improve communication accessibility and the societal integration of the deaf community in Portugal. Full article

The rise of the Internet of Things (IoT) has made it possible to explore different types of communication, such as underwater IoT (UIoT). This new paradigm allows the interconnection of ships, boats, coasts, objects in the sea, cameras, and animals that require constant monitoring. The use of sensors for environmental monitoring, tracking marine fauna and flora, and monitoring the health of aquifers requires the integration of heterogeneous technologies as well as wireless communication technologies. Aquatic mobile sensor nodes face various limitations, such as bandwidth, propagation distance, and data transmission delay issues. Owing to their versatility, wireless sensor networks support remote monitoring and surveillance. In this work, an architecture for a general packet radio service (GPRS) wireless sensor network is presented. The network is used to monitor the geographic position over the coastal area of the Gulf of Mexico. The proposed architecture integrates cellular technology and some ad hoc network configurations in a single device such that coverage is improved without significantly affecting the energy consumption, as shown in the results. The network coverage and energy consumption are evaluated by analyzing the attenuation in a proposed channel model and the autonomy of the electronic system, respectively. Full article

This paper presents an intelligent navigation wearable device for visually impaired individuals. The system aims to improve their independent travel capabilities and reduce the negative emotional impacts associated with visible disability indicators in travel tools. It employs an RGB-D camera and an inertial measurement unit (IMU) sensor to facilitate real-time obstacle detection and recognition via advanced point cloud processing and YOLO-based target recognition techniques. An integrated intelligent interaction module identifies the core obstacle from the detected obstacles and translates this information into multidimensional auxiliary guidance. Users receive haptic feedback to navigate obstacles, indicating directional turns and distances, while auditory prompts convey the identity and distance of obstacles, enhancing spatial awareness. The intuitive vibrational guidance significantly enhances safety during obstacle avoidance, and the voice instructions promote a better understanding of the surrounding environment. The device adopts an arm-mounted design, departing from the traditional cane structure that reinforces disability labeling and social stigma. This lightweight mechanical design prioritizes user comfort and mobility, making it more user-friendly than traditional stick-type aids. Experimental results demonstrate that this system outperforms traditional white canes and ultrasonic devices in reducing collision rates, particularly for mid-air obstacles, thereby significantly improving safety in dynamic environments. Furthermore, the system’s ability to vocalize obstacle identities and distances in advance enhances spatial perception and interaction with the environment. By eliminating the cane structure, this innovative wearable design effectively minimizes social stigma, empowering visually impaired individuals to travel independently with increased confidence, ultimately contributing to an improved quality of life. Full article