Learning Modality Complementary Features with Mixed Attention Mechanism for RGB-T Tracking
Yang Luo, Xiqing Guo, Mingtao Dong, Jin Yu
Abstract
RGB-T tracking involves the use of images from both visible and thermal modalities. The primary objective is to adaptively leverage the relatively dominant modality in varying conditions to achieve more robust tracking compared to single-modality tracking. An RGB-T tracker based on a mixed-attention mechanism to achieve a complementary fusion of modalities (referred to as MACFT) is proposed in this paper. In the feature extraction stage, we utilize different transformer backbone branches to extract specific and shared information from different modalities. By performing mixed-attention operations in the backbone to enable information interaction and self-enhancement between the template and search images, a robust feature representation is constructed that better understands the high-level semantic features of the target. Then, in the feature fusion stage, a modality shared-specific feature interaction structure was designed based on a mixed-attention mechanism, effectively suppressing low-quality modality noise while enhancing the information from the dominant modality. Evaluation on multiple RGB-T public datasets demonstrates that our proposed tracker outperforms other RGB-T trackers on general evaluation metrics while also being able to adapt to long-term tracking scenarios.