UniCOS

This is the official PyTorch codes for the paper.

Integrating Extra Modality Helps Segmentor Find Camouflaged Objects Well
Chengyu Fang, Chunming He, Longxiang Tang, Yuelin Zhang, Chenyang Zhu, Yuqi Shen, Chubin Chen, Guoxia Xu, Xiu Li, arXiv 2025

Abstract: Camouflaged Object Segmentation (COS) remains a challenging problem due to the subtle visual differences between camouflaged objects and backgrounds. Owing to the exceedingly limited visual cues available from visible spectrum, previous RGB single-modality approaches often struggle to achieve satisfactory results, prompting the exploration of multimodal data to enhance detection accuracy. In this work, we present UniCOS, a novel framework that effectively leverages diverse data modalities to improve segmentation performance. UniCOS comprises two key components: a multimodal segmentor, UniSEG, and a cross-modal knowledge learning module, UniLearner. UniSEG employs a state space fusion mechanism to integrate cross-modal features within a unified state space, enhancing contextual understanding and improving robustness to integration of heterogeneous data. Additionally, it includes a fusion-feedback mechanism that facilitate feature extraction. UniLearner exploits multimodal data unrelated to the COS task to improve the segmentation ability of the COS models by generating pseudo-modal content and cross-modal semantic associations. Extensive experiments demonstrate that UniSEG outperforms existing Multimodal COS (MCOS) segmentors, regardless of whether real or pseudo-multimodal COS data is available. Moreover, in scenarios where multimodal COS data is unavailable but multimodal non-COS data is accessible, UniLearner effectively exploits these data to enhance segmentation performance.

🏃 The architecture of the proposed UniCOS

🔥 News

• 2025-02-21: We release a part of results, bibtex, and the preprint of full paper.
• 2025-02-10: We release this repository, the preprint of full paper will be release soon.

🔧 Todo

• Complete this repository

🔗 Contents

• Datasets
• Training
• Testing
• Results
• Citation

🔍 Results

We achieved state-of-the-art performance on COD10K, CAMO, NC4K, CHAMELEON, and PCOD1200. More results can be found in the paper. We will release all results from different datasets when the paper is accepted.

Quantitative Comparison (click to expand)

• Quantitative results on UniCOS-Infrared and other methods.

  

• Quantitative results on UniCOS-Depth and other methods.

  

• Quantitative results on UniCOS-Polarization and other methods.

  

Visual Comparison (click to expand)

• Visual results on UniLearner

  

• Visual comparison on RGB and RGB-I

  

• Visual comparison on RGB-P and RGB-D

  

📎 Citation

If you find the code helpful in your research or work, please cite the following paper(s).

@misc{fang2025unicos,
      title={Integrating Extra Modality Helps Segmentor Find Camouflaged Objects Well}, 
      author={Chengyu Fang and Chunming He and Longxiang Tang and Yuelin Zhang and Chenyang Zhu and Yuqi Shen and Chubin Chen and Guoxia Xu and Xiu Li},
      year={2025},
      eprint={2502.14471},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2502.14471}, 
}