Multi-object Tracking via End-to-end Tracklet Searching and Ranking
arXiv:2003.02795
Abstract
Recent works in multiple object tracking use sequence model to calculate the similarity score between the detections and the previous tracklets. However, the forced exposure to ground-truth in the training stage leads to the training-inference discrepancy problem, i.e., exposure bias, where association error could accumulate in the inference and make the trajectories drift. In this paper, we propose a novel method for optimizing tracklet consistency, which directly takes the prediction errors into account by introducing an online, end-to-end tracklet search training process. Notably, our methods directly optimize the whole tracklet score instead of pairwise affinity. With sequence model as appearance encoders of tracklet, our tracker achieves remarkable performance gain from conventional tracklet association baseline. Our methods have also achieved state-of-the-art in MOT15~17 challenge benchmarks using public detection and online settings.
References in corpus (8)
- Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
- Simple Online and Realtime Tracking
- Deep Learning in Video Multi-Object Tracking: A Survey
- DenseBox: Unifying Landmark Localization with End to End Object Detection
- On Accurate Evaluation of GANs for Language Generation
- Tracklet Association Tracker: An End-to-End Learning-based Association Approach for Multi-Object Tracking
- FAMNet: Joint Learning of Feature, Affinity and Multi-dimensional Assignment for Online Multiple Object Tracking
- Evaluating Text GANs as Language Models