Multi-attention Associate Prediction Network for Visual Tracking

arXiv:2403.16395v1 Announce Type: new
Abstract: Classification-regression prediction networks have realized impressive success in several modern deep trackers. However, there is an inherent difference between classification and regression tasks, so they have diverse even opposite demands for feature matching. Existed models always ignore the key issue and only employ a unified matching block in two task branches, decaying the decision quality. Besides, these models also struggle with decision misalignment situation. In this paper, we propose a multi-attention associate prediction network (MAPNet) …

abstract arxiv attention block classification cs.cv difference diverse feature however issue key modern network networks prediction regression success tasks the key tracking type visual