Shifting Spotlight for Co-supervision: A Simple yet Efficient Single-branch Network to See Through Camouflage

2024-04-13

arXiv.org

AI Summary

To address the challenges of complex background interference and high computational overhead in camouflaged object detection, this paper proposes CS³Net—a single-branch network. Inspired by the biological phenomenon where moving light sources reveal camouflaged objects, we design a spotlight shifting strategy to generate boundary-aware supervision signals, eliminating redundancy inherent in multi-branch architectures. We introduce the first spotlight shifting collaborative supervision paradigm. Furthermore, we propose a Projection Aware Attention (PAA) module to enhance feature selectivity and an Extended Neighbor Connection Decoder (ENCD) to improve boundary modeling capability. Evaluated on standard benchmark datasets, CS³Net achieves state-of-the-art performance while reducing MACs by 32.13%, significantly advancing the accuracy-efficiency trade-off in camouflaged object detection.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Responsible Web: Measurement, analysis, and circumvention of Web censorship

Abstract

Camouflaged object detection (COD) remains a challenging task in computer vision. Existing methods often resort to additional branches for edge supervision, incurring substantial computational costs. To address this, we propose the Co-Supervised Spotlight Shifting Network (CS$^3$Net), a compact single-branch framework inspired by how shifting light source exposes camouflage. Our spotlight shifting strategy replaces multi-branch designs by generating supervisory signals that highlight boundary cues. Within CS$^3$Net, a Projection Aware Attention (PAA) module is devised to strengthen feature extraction, while the Extended Neighbor Connection Decoder (ENCD) enhances final predictions. Extensive experiments on public datasets demonstrate that CS$^3$Net not only achieves superior performance, but also reduces Multiply-Accumulate operations (MACs) by 32.13% compared to state-of-the-art COD methods, striking an optimal balance between efficiency and effectiveness.

Problem

• Object Detection
• Complex Backgrounds
• Computational Efficiency

Innovation

• CS$^3$Net
• PAA module
• ENCD module