SFGNet：语义与频率引导的网络用于伪装物体检测
摘要
伪装物体检测（Camouflaged Object Detection, COD）旨在分割融入环境中的物体。然而，大多数现有研究忽略了不同目标文本提示之间语义差异以及细粒度频率特征。本文提出了一种新型语义与频率引导网络（Semantic and Frequency Guided Network, SFGNet），其融合语义提示与频率域特征，以捕获伪装物体并提高边界感知。我们进一步设计了多带傅里叶模块（Multi-Band Fourier Module, MBFM）以增强网络处理复杂背景和模糊边界的能力。此外，我们设计了交互式结构增强模块（Interactive Structure Enhancement Block, ISEB），以确保预测结果的结构完整性与边界细节。对三个COD基准数据集上的广泛实验表明，我们的方法显著优于领先的技术方法。该模型的核心代码已提供链接：https://github.com/winter794444/SFGNetICASSP2026。
引用
@article{arxiv.2509.11539,
title = {SFGNet: Semantic and Frequency Guided Network for Camouflaged Object Detection},
author = {Dezhen Wang and Haixiang Zhao and Xiang Shen and Sheng Miao},
journal= {arXiv preprint arXiv:2509.11539},
year = {2025}
}
备注
Submitted to ICASSP 2026 by Dezhen Wang et al. Copyright 2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, including reprinting/republishing, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work. DOI will be added upon IEEE Xplore publication