Visual Sensing via Multiscale Edge-Aware Learning with Hybrid Attention for Camouflaged Object Detection

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 features undergo dual-path refinement through parallel $1 \times 1$ and $3 \times 3$ convolutions, preserving critical boundary information while enhancing semantic discriminability. Deeper features are recalibrated using Efficient Channel Attention modules with adaptive kernel selection. The refined multi-scale features are progressively fused and enhanced through an Edge Attention Module that explicitly strengthens boundary representations via gradient-based operations. Subsequently, an Attention over Attention mechanism performs hierarchical spatial refinement, enabling adaptive focus on discriminative regions while suppressing background distractions. Extensive experiments on four challenging benchmarks (CAMO, CHAMELEON, COD10K, NC4K) demonstrate that our framework achieves state-of-the-art performance.