Learning a Proposal Classifier for Multiple Object Tracking
arXiv:2103.07889
Abstract
The recent trend in multiple object tracking (MOT) is heading towards leveraging deep learning to boost the tracking performance. However, it is not trivial to solve the data-association problem in an end-to-end fashion. In this paper, we propose a novel proposal-based learnable framework, which models MOT as a proposal generation, proposal scoring and trajectory inference paradigm on an affinity graph. This framework is similar to the two-stage object detector Faster RCNN, and can solve the MOT problem in a data-driven way. For proposal generation, we propose an iterative graph clustering method to reduce the computational cost while maintaining the quality of the generated proposals. For proposal scoring, we deploy a trainable graph-convolutional-network (GCN) to learn the structural patterns of the generated proposals and rank them according to the estimated quality scores. For trajectory inference, a simple deoverlapping strategy is adopted to generate tracking output while complying with the constraints that no detection can be assigned to more than one track. We experimentally demonstrate that the proposed method achieves a clear performance improvement in both MOTA and IDF1 with respect to previous state-of-the-art on two public benchmarks. Our code is available at https://github.com/daip13/LPC_MOT.git.
Accepted at CVPR 2021, Poster, EEE/CVF Conference on Computer Vision and Pattern Recognition
References in corpus (18)
- Semi-Supervised Classification with Graph Convolutional Networks
- YOLOv3: An Incremental Improvement
- An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling
- Simple Online and Realtime Tracking
- MOT16: A Benchmark for Multi-Object Tracking
- Multivariate LSTM-FCNs for Time Series Classification
- Deep Learning in Video Multi-Object Tracking: A Survey
- MOT20: A benchmark for multi object tracking in crowded scenes
- Tracking the Trackers: An Analysis of the State of the Art in Multiple Object Tracking
- Deep Continuous Conditional Random Fields with Asymmetric Inter-object Constraints for Online Multi-object Tracking
- Lifted Disjoint Paths with Application in Multiple Object Tracking
- Spatial-Temporal Relation Networks for Multi-Object Tracking
- Simple Unsupervised Multi-Object Tracking
- Multiple People Tracking Using Hierarchical Deep Tracklet Re-identification
- GCNNMatch: Graph Convolutional Neural Networks for Multi-Object Tracking via Sinkhorn Normalization
- End-to-End Learning Deep CRF models for Multi-Object Tracking
- Fusion of Head and Full-Body Detectors for Multi-Object Tracking
- Multi-object Tracking via End-to-end Tracklet Searching and Ranking