High-Resolution Underwater Camouflaged Object Detection: GBU-UCOD Dataset and Topology-Aware and Frequency-Decoupled Networks

TL;DR

This paper introduces a high-resolution underwater camouflaged object detection dataset and a novel DeepTopo-Net framework that combines topology-aware modeling with frequency-decoupled perception to improve detection of slender and transparent marine creatures.

Contribution

The paper presents GBU-UCOD, the first high-resolution benchmark for marine vertical zonation, and proposes DeepTopo-Net with innovative modules for structural preservation and adaptive perception.

Findings

DeepTopo-Net outperforms existing methods on multiple datasets.

The GBU-UCOD dataset fills a critical data gap in deep-sea environments.

The proposed framework effectively preserves morphological features of underwater targets.

Abstract

Underwater Camouflaged Object Detection (UCOD) is a challenging task due to the extreme visual similarity between targets and backgrounds across varying marine depths. Existing methods often struggle with topological fragmentation of slender creatures in the deep sea and the subtle feature extraction of transparent organisms. In this paper, we propose DeepTopo-Net, a novel framework that integrates topology-aware modeling with frequency-decoupled perception. To address physical degradation, we design the Water-Conditioned Adaptive Perceptor (WCAP), which employs Riemannian metric tensors to dynamically deform convolutional sampling fields. Furthermore, the Abyssal-Topology Refinement Module (ATRM) is developed to maintain the structural connectivity of spindly targets through skeletal priors. Specifically, we first introduce GBU-UCOD, the first high-resolution (2K) benchmark tailored for marine vertical zonation, filling the data gap for hadal and abyssal zones. Extensive experiments on MAS3K, RMAS, and our proposed GBU-UCOD datasets demonstrate that DeepTopo-Net achieves state-of-the-art performance, particularly in preserving the morphological integrity of complex underwater patterns. The datasets and codes will be released at https://github.com/Wuwenji18/GBU-UCOD.