State-aware Re-identification Feature for Multi-target Multi-camera Tracking
This work addresses tracking accuracy issues in multi-camera surveillance systems, representing an incremental improvement by enhancing re-identification features with state-aware information.
The paper tackled the problem of unreliable appearance features in Multi-target Multi-camera Tracking (MTMCT) due to occlusion and orientation variance, which cause identity switches and tracklet fragments, by proposing a novel tracking framework that incorporates occlusion status, orientation, and human pose information into the re-identification model, achieving 81.3% IDF1 on a hard sequence and outperforming other methods.
Multi-target Multi-camera Tracking (MTMCT) aims to extract the trajectories from videos captured by a set of cameras. Recently, the tracking performance of MTMCT is significantly enhanced with the employment of re-identification (Re-ID) model. However, the appearance feature usually becomes unreliable due to the occlusion and orientation variance of the targets. Directly applying Re-ID model in MTMCT will encounter the problem of identity switches (IDS) and tracklet fragment caused by occlusion. To solve these problems, we propose a novel tracking framework in this paper. In this framework, the occlusion status and orientation information are utilized in Re-ID model with human pose information considered. In addition, the tracklet association using the proposed fused tracking feature is adopted to handle the fragment problem. The proposed tracker achieves 81.3\% IDF1 on the multiple-camera hard sequence, which outperforms all other reference methods by a large margin.