Joint Object Detection and Multi-Object Tracking with Graph Neural Networks

23 Jun 2020  ·  Yongxin Wang, Kris Kitani, Xinshuo Weng ·

Object detection and data association are critical components in multi-object tracking (MOT) systems. Despite the fact that the two components are dependent on each other, prior works often design detection and data association modules separately which are trained with separate objectives. As a result, one cannot back-propagate the gradients and optimize the entire MOT system, which leads to sub-optimal performance. To address this issue, recent works simultaneously optimize detection and data association modules under a joint MOT framework, which has shown improved performance in both modules. In this work, we propose a new instance of joint MOT approach based on Graph Neural Networks (GNNs). The key idea is that GNNs can model relations between variable-sized objects in both the spatial and temporal domains, which is essential for learning discriminative features for detection and data association. Through extensive experiments on the MOT15/16/17/20 datasets, we demonstrate the effectiveness of our GNN-based joint MOT approach and show state-of-the-art performance for both detection and MOT tasks. Our code is available at: https://github.com/yongxinw/GSDT

PDF Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Multi-Object Tracking 2D MOT 2015 GSDT MOTA 60.7 # 1
Multi-Object Tracking MOT16 GSDT MOTA 66.7 # 13
Multi-Object Tracking MOT17 GSDT MOTA 66.2 # 33
Multi-Object Tracking MOT20 GSDT MOTA 67.1 # 17

Methods


No methods listed for this paper. Add relevant methods here