Recent advances in unsupervised video object segmentation have highlighted the potential of two-stream architectures that integrate appearance and motion cues. However, fully leveraging these complementary sources of information requires effectively modeling their interdependencies. In this paper, we introduce cross-modality token modulation, a novel approach designed to strengthen the interaction between appearance and motion cues. Our method establishes dense connections between tokens from each modality, enabling efficient intra-modal and inter-modal information propagation through relation transformer blocks. To improve learning efficiency, we incorporate a token masking strategy that addresses the limitations of relying solely on increased model complexity. Our approach achieves state-of-the-art performance across all public benchmarks, outperforming existing methods.
@article{arxiv.2604.14630,
title = {CMTM: Cross-Modal Token Modulation for Unsupervised Video Object Segmentation},
author = {Inseok Jeon and Suhwan Cho and Minhyeok Lee and Seunghoon Lee and Minseok Kang and Jungho Lee and Chaewon Park and Donghyeong Kim and Sangyoun Lee},
journal= {arXiv preprint arXiv:2604.14630},
year = {2026}
}