An estimated ten million people live with Parkinson's, yet timely diagnosis remains constrained by specialist assessment, costly imaging, and unequal care access. Phonation analysis offers a low-cost alternative: dopaminergic degeneration causes vocal impairments years before motor symptoms emerge, enabling community screening. However, existing ML approaches are limited by recording-level data leakage and opacity. This paper presents a four-phase acoustic framework-spanning signal acquisition, feature processing, and interpretable decision support-applied to a multi-type Parkinson's speech dataset (40 training, 28 blind-test). The framework enforces subject-level partitioning via GroupKFold... More >

Deepfake technology has advanced swiftly, enabling the rapid production of hyper-realistic synthetic media that pose considerable threats to digital security, privacy, military operations, and information integrity. This paper extensively examines visual intelligence and computer vision methodologies for deepfake detection, covering recent developments in deep learning, adversarial strategies, and feature extraction. It reviews prevalent generation architectures—including GANs, autoencoders, neural rendering, and diffusion models—alongside novel adversarial tactics that enhance realism while evading detection, particularly in military and intelligence contexts. We also investigate visual... More >

Surface defect segmentation in metallic materials presents significant challenges due to irregular defect shapes, extreme scale variations ranging from small blowholes to large uneven regions, and low contrast with complex background textures. While Convolutional Neural Networks excel at local feature extraction, their limited receptive fields hinder effective global context modeling. Conversely, Vision Transformers capture long-range dependencies but struggle with fine-grained boundary details critical for accurate defect localization. To address these limitations, we propose FocusNet, a novel architecture integrating multi-scale feature refinement, hybrid attention mechanisms, and progress... More >

Cognitive-semantic communication encodes task-relevant meaning rather than raw bits, improving 6G spectral efficiency and reducing latency, yet evidence remains fragmented across methods, evaluations, and deployments. This PRISMA-based review screens 782 records (2019-2025) from seven databases, retaining 73 studies organized into five method families: neural semantic codecs, KG-based pipelines, LLM-assisted codecs, cross-modal encoders, and semantic relays. We examine how goal-oriented semantics, knowledge graphs, and LLMs shape representation, compression, and task quality under rate and latency constraints across UAVs, NTNs, ISAC, Metaverse, and IoT. Reported gains include bandwidth reduc... More >

Salient object detection aims to identify and segment the most visually prominent objects in images. Despite significant advances in deep learning, existing methods struggle to balance global context modeling, boundary preservation, and multi-scale feature integration. To address these limitations, we propose MAFNet (Multi-level Attention Fusion Network), a novel attention-driven framework that leverages specialized attention mechanisms tailored to different semantic levels. Our approach employs a Tokens-to-Token (T2T) Transformer backbone for hierarchical feature extraction, capturing both local structural details and global contextual relationships. The core contribution lies in a comprehe... More >

Precise segmentation of colorectal polyps in colonoscopy images is essential for timely cancer diagnosis and prevention. Nevertheless, current segmentation methods contend with intrinsic variability in polyp appearance, differences in size, shape, and texture, while preserving computational efficiency necessary for clinical implementation. In this paper, we present a novel segmentation architecture that integrates scale-specific attention mechanisms within a hybrid CNN-Transformer backbone to address these challenges. Our model employs Coordinate Attention for high-resolution feature maps to preserve spatial details essential for boundary delineation, and Channel Attention for deep semantic... More >

Falls are a major cause of injury, hospitalization, and loss of independence among older adults, spurring interest in visual intelligence-based automated fall detection for timely response and continuous monitoring. This article presents a systematic review of such systems, focusing on YOLO-based approaches. Following PRISMA guidelines, the review covers 2016–2025 literature, identifying 637 records and including 63 studies after screening. We examine datasets, preprocessing strategies, evaluation protocols, metrics, and hardware platforms, comparing reported accuracy, efficiency, and real-time feasibility across different designs. Evidence is strongest for YOLOv3 through YOLOv9, while evi... More >

Camouflaged object detection (COD) remains a significant challenge in computer vision. Existing approaches struggle to address both body immersion and structural ambiguity simultaneously, leading to inaccurate boundary delineations. This paper presents a novel Visual Sensing framework via Multiscale Edge-Aware Learning with Hybrid Attention. The proposed framework integrates hierarchical feature extraction, adaptive attention mechanisms, and progressive multi-scale fusion to achieve robust COD. We employ EfficientNetB7 as the backbone network to extract six-scale hierarchical features, capturing both fine-grained spatial details and high-level semantic representations. Initial shallow featur... More >