Globally Consistent Multi-People Tracking using Motion Patterns
This addresses the challenge of more accurate and reliable tracking in video surveillance or crowd analysis, though it is incremental as it builds on and enhances current methods.
The paper tackles the problem of multi-people tracking by imposing global consistency constraints on trajectories, which are typically missing in state-of-the-art methods that rely on local smoothness. The result is improved performance when used with existing algorithms, including an unsupervised scheme that achieves similar gains without ground truth.
Many state-of-the-art approaches to people tracking rely on detecting them in each frame independently, grouping detections into short but reliable trajectory segments, and then further grouping them into full trajectories. This grouping typically relies on imposing local smoothness constraints but almost never on enforcing more global constraints on the trajectories. In this paper, we propose an approach to imposing global consistency by first inferring behavioral patterns from the ground truth and then using them to guide the tracking algorithm. When used in conjunction with several state-of-the-art algorithms, this further increases their already good performance. Furthermore, we propose an unsupervised scheme that yields almost similar improvements without the need for ground truth.