Search Results (3,825)

Egocentric multi-view image analysis refers to the processing of utilizing synchronized video streams captured from multiple wearable cameras worn on the head or body, providing complementary first-person perspectives of dynamic, real-world interactions. Unlike single-view egocentric vision, which may suffer from severe occlusions, motion blur, and limited field-of-view or traditional fixed-camera multi-view setups (assuming static geometry and controlled environments), egocentric multi-view systems leverage body-worn rigs to enable a more robust and flexible 3D understanding in open-world, mobile scenarios. In this work, we present a systematic survey of advancements in cross-view feature fusion, geometric consistency enforcement, open-world detection, human–object interaction (HOI) modeling, action segmentation, 3D reconstruction, and novel-view synthesis specifically tailored to wearable multi-camera platforms. Key datasets released between 2024 and 2026—including HOT3D (833 min of synchronized multi-view hand/object interactions from Project Aria and Quest 3), MultiEgo (first multi-egocentric dataset for 4D social scene reconstruction), and Ego-1K (large-scale 12-camera rig for dynamic 3D video synthesis) are thoroughly examined alongside an analysis of integrations with large language models (LLMs) and vision–language models that drive performance gains, typically in the 15–30% range over single-view baselines in hand tracking, HOI recognition, and reconstruction fidelity, although we show through a consolidated meta-analysis that this gain is task-dependent: larger for geometry-bottlenecked tasks such as in-hand object lifting, and smaller, method-dependent, or occasionally negative for semantic-recognition tasks such as keystep recognition under naive view fusion. These methods cover work in multi-view stereo, cross-view learning, and novel-view synthesis while addressing several real-time wearable constraints. Practical applications such as immersive Augmented Reality/Virtual Reality (AR/VR), assistive robotics, and healthcare monitoring are also discussed together with the challenges in motion calibration, benchmark diversity, and edge deployment ability. Thus, in this review, we attempt to fill a critical gap by focusing exclusively on wearable multi-view systems in an open-world setting, synthesizing the latest literature to chart future directions toward more embodied and continual learning agents. Full article

Background: Anemia is a common systemic condition associated with adverse maternal, perioperative, and cardiovascular outcomes. Although timely screening is clinically important, diagnosis still relies on invasive blood testing. Palpebral conjunctival pallor has traditionally been used as a noninvasive indicator of anemia, but its diagnostic accuracy remains limited. This study aimed to develop and validate a deep learning system to estimate hemoglobin (Hb) concentration and screen for anemia using palpebral conjunctiva images captured with a smartphone-compatible slit-lamp microscope. Methods: In this prospective observational study, 225 Japanese participants (20–92 years) underwent conjunctival imaging and blood testing. Palpebral conjunctiva videos were obtained using the Smart Eye Camera. Video frames were processed using automated anterior-segment segmentation and conjunctiva extraction. A ConvNeXt-based regression model was trained to predict Hb values. Anemia was defined using sex-specific Hb thresholds. Results: From 225 videos, 53,776 frames were extracted, yielding 9903 quality-filtered conjunctiva images (training: 8082; test: 1821). Video-level predicted Hb values moderately correlated with measured Hb (r = 0.42). For anemia screening, frame-level analysis achieved an AUC of 0.75, with accuracy of 0.76, sensitivity of 0.71, and specificity of 0.79. Video-level aggregation achieved 69% accuracy. Conclusions: Deep learning analysis of palpebral conjunctiva images acquired with a portable slit-lamp microscope demonstrated the feasibility of non-invasive hemoglobin estimation and anemia screening. Although the proposed approach achieved moderate performance, further improvements in model accuracy and prospective multi-center validation are required before clinical implementation. Full article

Small shelter structures may affect the spatial use, locomotor state, and group spatial relationships of juvenile black seabream (Acanthopagrus schlegelii), but behavioral responses induced by different structure types remain insufficiently examined using multiple behavioral metrics. In this study, we conducted a tank experiment to test three hypotheses: compared with a structure-free control, cuboid and cylindrical structures would increase juvenile use of the central structure-placement zone, alter locomotor state, and affect inter-individual distance. Three treatments were established, including cuboid structure, cylindrical structure, and structure-free control, and fish behavior was observed under daytime and nighttime conditions. Video tracking and image analysis were used to quantify area-standardized average occurrence rate, average swimming speed, percentage of active time, and inter-individual distance. Both structures increased juvenile use of S1, namely the central structure-placement zone within a radius of 0–20 cm from the shelter-structure center. During the daytime, the average occurrence rates in S1 were 54.1% ± 6.33% and 46.7% ± 2.32% in the cuboid and cylindrical structure groups, respectively, which were higher than those in the structure-free control group (15.2% ± 3.08%). At night, the corresponding values were 50.2% ± 0.55%, 42.8% ± 3.77%, and 13.3% ± 1.57%, respectively. Locomotor behavior and inter-individual distance also varied with structure type, observation period, and diel condition. Under tank conditions, small shelter-like structures altered the short-term spatial use, locomotor state, and group spatial relationships of juvenile black seabream. These findings provide experimental evidence for how juvenile black seabream respond to simple shelter-like structures under controlled tank conditions and may help researchers, hatchery managers, and stock-enhancement practitioners select small structures for short-term behavioral assessment and shelter enrichment. Full article

Cyberattacks via email remain a major menace to people, companies, and critical infrastructures, and effective spam and phishing detection is a social concern. Nevertheless, the current methods, such as NetSpam and SMART, tend to have issues with non-homogenous data streams, lack of contextual knowledge, poor generalization, and inability to adapt to changing attack patterns. The existing techniques are not strong in terms of multimodal fusion and cannot effectively transfer trust or update risk scores in dynamic conditions. To overcome these shortcomings, this paper presents a Zone-Aware Multimodal Graph-Based Transformer that combines text, image, video, and metadata streams in a smooth manner to detect threats in emails. The three key novelties of the proposed framework include AAGFusion to contrastively align multimodal features and hierarchically fuse them using transformers; MAGNN-SASO to classify zones, compute similarity across zones, and optimize bio-inspired optimization; and Q-BayesTrustNet-X to propagate trust, risk score, Bayesian calibration, and continual learning, and provide interpretable feedback by using LRP-based explainability. The experimental findings prove that the proposed system has a high level of performance, with the accuracy, precision, and specificity reaching 98.57, 97.51, and 99.48, respectively, which proves its efficiency in high-fidelity and real-world spam and phishing detection. Full article

Reliable three-dimensional (3D) reconstruction from endoscopic video is essential for endoscopic digital twins, scene review, and minimally invasive visual analysis. However, endoscopic images are not clean observations of intrinsic tissue appearance. Depth-dependent blur, shallow mucosal color diffusion, wet-surface specular reflection, and frame-wise color variation are often coupled with the captured signal. When such observation-dependent effects are directly optimized as Gaussian colors, conventional 3D Gaussian Splatting may encode transient imaging artifacts as persistent tissue appearance, leading to blurred textures, color drift, specular residues, and unstable novel-view synthesis. This paper presents EndoDGS (Endoscopic Degradation-Decoupled Gaussian Splatting), a degradation-decoupled Gaussian Splatting framework for endoscopic novel-view reconstruction. The core idea is to keep stable geometry and base tissue appearance in the Gaussian representation, while modeling endoscope-induced degradations separately in a bounded render-space compensation pipeline. EndoDGS combines lightweight appearance modulation for frame-wise color stabilization with sequential degradation compensation for optical blur, mucosal color transport, and wet-surface specular response. This design reduces the entanglement between persistent tissue appearance and transient imaging degradations without changing the underlying Gaussian geometry and visibility ordering. Experiments on synthetic colonoscopy and real endoscopic/laparoscopic datasets covering 38 scenes show that EndoDGS consistently improves reconstruction quality over representative implicit and explicit reconstruction baselines. The results demonstrate that separating stable tissue representation from observation-dependent endoscopic degradations provides a more faithful, stable, and interpretable foundation for endoscopic 3D reconstruction. Full article

This paper presents the Micro Video Guidance Sensor Version 2 (uVGS-2), a ROS-based vision navigation framework for real-time six-degrees-of-freedom pose estimation in drones, spacecraft, and autonomous robotic platforms operating in GNSS-denied environments. The system evolves from the previous Smartphone Video Guidance Sensor (SVGS) architecture through a modular C++ implementation, including advanced image preprocessing, deterministic blob sorting, and an optimized perspective-4-point solver using a Lie-algebra-based analytical Jacobian formulation. The proposed architecture achieves computationally efficient photogrammetric state estimation using onboard camera and processor resources, enabling deployment in resource-constrained systems. Experimental validation was conducted in NASA’s Astrobee free-flying robot, both at the International Space Station (ISS), for SVGS, and by ground testing through real-time sensor-fusion with Astrobee’s graph-based localizer (Astroloc), for uVGS-2. Results demonstrate robust centimeter-level accuracy in relative position and attitude estimation under illumination disturbances, partial occlusions, and intermittent loss of line-of-sight. The framework can be used in robotic platforms and autonomous UAV operations, including precision landing, formation flight, and cooperative navigation in environments where GNSS signals are unavailable or intermittent. Full article

This study aimed to quantitatively analyze giant slalom turn kinematics using a video-based three-dimensional computer-aided design (3D CAD) model-matching method. The analyzed skiers included the top five finishers (WT) and one top-ranked Japanese elite skier (JP). Broadcast competition footage was analyzed to estimate the ski edging angle, center of mass (COM) position, and knee joint and hip joint angle during the turning motion. One turn cycle was divided into five phases: neutral position (NP), edging start (ES), load start (LS), turn maximum (TM), and edging finish (EF). Phase durations and kinematic variables were compared descriptively. JP demonstrated a longer total turn duration than WT, particularly during the LS–TM phase. During the NP–ES phase, WT elevated the COM while simultaneously shifting it inward, whereas JP lowered the COM. JP demonstrated greater hip and knee flexion during the early turn phase compared with WT. This study demonstrates the feasibility of extracting quantitative kinematic information from broadcast competition footage using a 3D CAD model-matching approach. The findings suggest that differences in COM and lower-limb movements may reflect different turning movement strategies. Because only one Japanese elite skier was analyzed, the findings should be interpreted as exploratory observations and cannot be generalized to all Japanese alpine skiers. Full article

The performance of Network-on-Chip (NoC) architectures is severely limited by the high dynamic power consumption and interconnect crosstalk with the scaling of deep sub-micron (DSM) technology. The coupling capacitance between wires dominates the self-capacitance in dense interconnects, and the coupling transitions contribute the most to the link power dissipation. In this paper, we propose a new low-power error correction coding scheme called Parity-based Multi-bit Transient Error Correction (PMTEC), which can simultaneously deal with self-transitions and coupling transitions in NoC links. The proposed method reduces the effective switching activity by employing a balanced encoding strategy, unlike existing coding techniques, which typically degrade the signal integrity or impose significant power overheads for particular data patterns. Experimental analysis demonstrates that PMTEC achieves strong link power reductions on various data profiles, including text, image, and video traffic, without the negative overheads of cutting-edge techniques like Duplicated Two-Dimensional Parities (DTDPs) and Crosstalk-Aware Transient Error Correction (CATEC). Furthermore, the codec’s parametric synthesis reveals a very small footprint, taking up only 2472.018 $\mathsf{\mu }{\mathrm{m}}^2$ of silicon area and using a negligible 0.219 $\mathsf{\mu }$W of power, which are 87.57% and 99.20% reductions over CATEC, respectively. The suggested work offers a scalable and energy-efficient solution for dependable on-chip communication in power-constrained systems, despite requiring a 15.371 ns latency trade-off to correct 8-bit bursts. Full article

This paper presents a lightweight MCU–EdgeTPU platform—a microcontroller unit (MCU) paired with an Edge Tensor Processing Unit (EdgeTPU) accelerator—for onboard drone-perception experiments, extended from an open-source baseline originally limited to Quarter Video Graphics Array (QVGA) single-camera operation. Rather than treating hardware, runtime, model, and data as separate problems, they are developed as parts of the same continuous perception pipeline. The platform extends the hardware baseline toward dual 5 Mpx sensing, onboard inertial measurement unit (IMU) support, real-time embedded inference, and a high-level MicroPython control layer. In parallel, lightweight You Only Look Once (YOLO) detectors are trained and selected on a synthetic aerial-person dataset generated under the visual conditions expected by the drone camera, including target resolution, viewpoint, object scale, weather, lighting, and time-of-day variation. The resulting workflow starts from both ends: the detector must be small and quantization-stable enough for the EdgeTPU path, while the dataset must match the images that the onboard sensor is expected to observe. To evaluate the system, the full path from camera capture and image conversion to TPU transfer, model execution, and post-inference processing is analyzed. In the tested setup, the optimized single-camera pipeline runs stably with no timeouts or inference failures at about 26 detections per second with standard RGB input; because each EdgeTPU invocation is bounded by the USB transfer of the input image, feeding the camera’s native YUV420 format instead halves that transfer and raises throughput to about 40 detections per second at the same accuracy, while the selected 8-bit-integer (INT8) person detector preserves most of its 32-bit floating-point (FP32) accuracy. Detections are exposed to drone-control workflows (MAVLink/PX4 and Crazyflie) through the scriptable layer as an integration interface rather than a validated autonomy stack. The central contribution is therefore a co-designed embedded perception pipeline in which the board, runtime, detector, dataset, and even the camera pixel format are aligned around the same operating conditions. Full article

To address the challenges of the large-scale georeferencing of tower-based cameras and the limited capability of video-based spatial analysis, we proposed a geospatial mapping method integrating 3D GIS and gradient descent optimization. Using a Digital Elevation Model (DEM), high-resolution remote sensing imagery, and tower-based video data as the primary data sources, the proposed method first estimates the intrinsic parameters of the tower-based camera by aligning a 3D GIS virtual camera with the video imagery. Subsequently, the initial camera extrinsic parameters are estimated using the PnP algorithm based on the previously estimated intrinsic matrix $K$ and the corresponding control point pairs. Building upon these initial estimates, the camera intrinsic and extrinsic parameters are jointly optimized using a constrained L-BFGS-B framework that incorporates prior knowledge of the tower planar location, explicit box constraints, and a semi-constrained parameterization scheme with bounded parameter ranges. Furthermore, an outlier-removal and re-optimization strategy is employed to further improve the accuracy of parameter estimation. Finally, the optimized parameters are employed to transform image coordinates into three-dimensional world coordinates, and video geospatial mapping is achieved through the integration of colored point clouds with the 3D GIS scene. The results showed the following: (1) The 3D GIS scene constructed from publicly available DEM and high-resolution remote sensing imagery met the requirements for the initial estimation of intrinsic and extrinsic camera parameters. (2) Compared with PnP, RANSAC-PnP, SQPnP, and DLT, the proposed method achieves lower reprojection and 3D spatial errors. For the independent check points, the RMSE of the reprojection error is reduced by 66.4%, 73.6%, 68.0%, and 48.3%, respectively, while the RMSE of the 3D spatial error is reduced by 84.6%, 86.2%, 83.1%, and 69.4%, respectively. These results demonstrate that the proposed method provides reliable camera parameter estimates for video geospatial mapping. (3) Using the estimated camera parameters, image coordinates are transformed into 3D world coordinates to generate a georeferenced colored point cloud, which facilitates integrated analysis with existing geospatial datasets. The proposed method provides a feasible solution for tower-based camera georeferencing and three-dimensional visualization under conditions without field calibration. It offers a theoretical and technical basis for geospatial monitoring and related applications. Full article

This qualitative media-archaeology study examines how digital video compression can reshape viewer perception of a legacy television image. A low-resolution online image from the 1976 The Mary Tyler Moore Show is compared with the corresponding image from a commercial DVD. In the compressed 360p online version, Moore’s upper torso appears briefly overexposed during a bubble-bath scene—an interpretation often described as a wardrobe malfunction. However, the higher-resolution DVD clearly shows that Moore maintained broadcast standards by wearing a protective undergarment that became visually erased in the compressed media. This divergence serves as a result of a natural experiment, demonstrating how low-resolution encoding of an image produces edge smoothing, tonal blending, and dissolution of material boundaries. These transformations support an inductive interpretation of materiality collapse, a compression artifact in which garments, skin, and shadows lose visual distinctiveness—creating an image of Mary Tyler Moore perceptually similar to a classical nude sculpture such as the Venus de Milo. Contextual evidence from Moore’s autobiography further clarifies production norms that shaped the bubble-bath scene and contributed to insufficient foam coverage. The findings show how compressed digital video can generate culturally consequential misperceptions, underscoring the need to scrutinize online compressed images posted as material evidence. Full article

When subjected to compound interference, such as spatial noise and high-frequency jitter, current LGMD-inspired collision detection models for micro-robots are prone to false alarms and perceptual degradation. To address this challenge, this paper proposes an enhanced visual perception model that incorporates adaptive Gaussian random variables and spatial residual feedback (SRF). These random variables filter out discrete spatial noise, while the SRF suppresses global image shifts induced by jitter. Evaluations on synthetic and real-world video sequences validate the proposed mechanisms. Comparative results demonstrate that the model effectively reduces false responses under compound interference, thereby maintaining robust success rate (SR), discrimination ratio (DR), and membrane potential stability index (MPSI) metrics. To explain this robustness, ablation analyses further verify the synergistic benefits of the SRF and the Gaussian random variables. Furthermore, statistical results on the random variables indicate that, under compound interference, the adaptive probabilistic model outperforms fixed probabilistic configurations. By ensuring robust collision perception against such interference, this work enhances the practical viability of LGMD-inspired visual systems. Full article

The Upper Niger National Park (UNNP) is the oldest park and one of the most promising conservation areas in Guinea; yet, its mammalian fauna remains poorly documented. Camera trapping has become an essential tool for revealing the diversity and assemblage structure of such communities. This study employed camera traps to improve knowledge about terrestrial and semi-terrestrial mammals with body mass > 0.5 kg (Sciuridae and heavier) in the Mafou Fully Protected Area (Mafou FPA), the principal core zone of the UNNP. The survey was conducted during the dry season, from January to May 2025. Sampling targeted animal trails within forest habitats in the Mafou FPA and involved the deployment of 53 camera traps with an average inter-trap distance of 2 km. Across 4239 camera-days of sampling effort, we collected 10,334 usable images and videos, yielding 2634 independent detection events. Thirty taxa across 15 families and five orders, including six species of high conservation concern according to the IUCN Red List (version 2025-1) were recorded. The mammal assemblage consists of species from the three known trophic levels (prey, mesopredators, and apex predators), with predominance of medium-sized prey species. Our results demonstrate a remarkable richness in frugivorous seed-dispersing species, ecosystem-engineering species, and the presence of a megaherbivore, which all contribute to its ecological dynamics. Despite a notable human activity index of 0.42, occupancy models revealed broad spatial distribution of medium-sized mammals within this protected area. The occupancy patterns demonstrated that occurrence of some species might be sensitive to human disturbances in the area. These findings are an important contribution to Guinean and West African biodiversity assessments. They highlight the critical conservation value of the UNNP within the subregion. Full article

Galloping of overhead transmission lines threatens grid safety and requires non-contact measurement methods that can quantify three-dimensional (3D) motion from field video. This paper proposes a deep-vision and binocular-collaboration framework for dynamic conductor displacement measurement. The framework combines three components that are matched to the physical structure of transmission lines: adaptive image enhancement using Retinex illumination decomposition and Wiener blind deconvolution; a structure-prior dual-branch extraction module that uses an improved YOLOv11 keypoint branch for spacer-equipped sections and an improved U-Net branch with Dynamic Snake Convolution (DSC) and Strip Pooling for bare conductors; and stereo reconstruction with Kalman-filter-based temporal association for continuous trajectory estimation. Compared with the original submission, the revised manuscript further clarifies the real-video data acquisition, annotation procedure, camera synchronization, calibration workflow, training/testing independence, and runtime measurement protocol. Additional validation on a public real power-line image dataset is also reported. The proposed method achieves a Z-axis Root Mean Square Error (RMSE) of 24.5 mm for spacer sections in the controlled binocular field test, a dominant-frequency relative error below 3.5%, and 32 FPS on edge hardware when preprocessing, visual extraction, stereo projection, and temporal filtering are included. On the supplementary public power-line dataset, the segmentation branch obtains a Dice coefficient of 0.9039 and an IoU of 0.8395. These results indicate that the proposed framework reduces the depth-scale limitation of monocular vision and provides a practical quantitative tool for field galloping monitoring. Full article

Vehicle-ejected debris detection is a practical but insufficiently studied problem in intelligent traffic enforcement. Unlike static road litter, objects thrown from moving vehicles are usually small, irregular, transient, and easily confused with road textures, shadows, lane markings, and light reflections. In current traffic management, such violations still rely heavily on manual video review or offline inspection, while task-specific datasets and edge-deployable detection solutions remain limited. To address this gap, this study constructs a vehicle-ejected debris dataset containing 4328 annotated image samples collected from real road scenarios. The dataset covers urban and suburban roads, daytime and nighttime illumination, near-range and distant small-object cases, and hard negative samples. To meet the coupled requirements of vehicle-mounted small-object detection and edge-side INT8 deployment, this study develops a hardware-aware lightweight detection framework based on YOLOv8m. The original CSPDarknet backbone is replaced with the convolutional variant of MobileNetV4 to reduce feature-extraction cost, while a scale-specific Channel Alignment Module is inserted between the heterogeneous MobileNetV4 backbone and the YOLOv8m PANet neck to preserve multi-scale feature compatibility. The alignment module uses only BPU-friendly convolution, batch normalization, and activation operations, thereby avoiding deployment-unfriendly operators while maintaining compatibility with INT8 quantization and edge acceleration. The trained FP32 model is quantized to INT8 and deployed on the RDK X5 BPU using the Horizon OpenExplorer toolkit. Experimental results and repeated-seed validation show that the proposed model achieves a consistent accuracy–efficiency advantage on the constructed dataset. In a representative run, the proposed model obtains 93.1% mAP₅₀, while reducing the number of parameters from 25.9 M to 13.1 M and GFLOPs from 78.9 to 39.6 compared with the YOLOv8m baseline. After INT8 deployment, the model reaches 112.6 FPS on the RDK X5 platform with only a minor accuracy decrease. These results indicate that the proposed framework can serve as a practical edge-deployable perception module for real-time vehicle-ejected debris monitoring under vehicle-mounted traffic-enforcement scenarios. It should be noted that this work focuses on single-frame debris detection, while event-level ejection verification, temporal consistency analysis, offending-vehicle attribution, and enforcement decision-making remain beyond the scope of this study. Full article