Abstract
Multi-object tracking is a crucial problem for autonomous vehicle. Most state-of-the-art approaches adopt the tracking-by-detection strategy, which is a two-step procedure consisting of the detection module and the tracking module. In this paper, we improve both steps. We improve the detection module by incorporating the temporal information, which is beneficial for detecting small objects. For the tracking module, we propose a novel compressed deep Convolutional Neural Network (CNN) feature based Correlation Filter tracker. By carefully integrating these two modules, the proposed multi-object tracking approach has the ability of re-identification (ReID) once the tracked object gets lost. Extensive experiments were performed on the KITTI and MOT2015 tracking benchmarks. Results indicate that our approach outperforms most state-of-the-art tracking approaches.
Highlights
Multi-object tracking is one of the most fundamental capabilities for the autonomous vehicle.The autonomous vehicle mostly works in the dynamic environment
Our approach achieved the best result in Multiple Object Tracking Accuracy (MOTA), Mostly Track targets (MT) and Mostly Lost targets (ML)
MDP [12] achieved a better result in ID F1 score (IDF1), while it performed much worse in MT and ML
Summary
Multi-object tracking is one of the most fundamental capabilities for the autonomous vehicle. The autonomous vehicle mostly works in the dynamic environment. Through precisely tracking other dynamic objects’ movements, the autonomous vehicle could plan its own trajectory and run smoothly. Existing multi-object tracking approaches mostly adopt the tracking-by-detection strategy. The potential objects-of-interest are detected using object detection algorithm. These potential objects are linked across different frames to form the so-called tracklets in the second step
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call