Multiple object tracking using a dual-attention network for autonomous driving

Multiple object tracking (MOT) remains an open and challenging problem for autonomous vehicles. Existing methods mainly ignore prior information from real traffic scenes. Here, the authors propose a novel MOT algorithm that considers traffic safety for vulnerable road users. The proposed method integrates two attention modules with a novel detection refinement strategy. Since skilled drivers pay more attention to pedestrians and cyclists, the authors employ a saliency detection method to extract scene attention region. Then, a detection refinement strategy achieved a good trade-off between parallel single object trackers and detection results. Channel attention can mine the most useful feature channel for traffic road users. In the end, the authors operate their method on the popular MOT 17 benchmark in comparison with other high-level MOT algorithms. The tracking results show that the proposed dual-attention network achieves the state-of-the-art performance.