Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

UCMCTrack: Multi-Object Tracking with Uniform Camera Motion Compensation

Kefu Yi, Kai Luo, Xiaolei Luo, Jiangui Huang, Hao Wu, Rongdong Hu, Wei Hao

TL;DR

This paper addresses multi-object tracking under challenging camera motion by designing UCMCTrack, a pure motion-model tracker that operates on the ground plane. It replaces frame-by-frame camera motion compensation with a single sequence-wide parameter set and adopts a Kalman filter on ground-plane states, using the Mapped Mahalanobis Distance to perform data association with ground-plane uncertainty modeled explicitly. Through correlated ground-plane measurements and a process noise compensation scheme, UCMCTrack achieves state-of-the-art results on MOT17, MOT20, DanceTrack, and KITTI, while maintaining real-time efficiency (>$1000$ FPS on CPU). The work highlights the benefit of grounding motion modeling in the ground plane and suggests future improvements by integrating traditional appearance cues (IoU, ReID) to further enhance robustness and generalization.

Abstract

Multi-object tracking (MOT) in video sequences remains a challenging task, especially in scenarios with significant camera movements. This is because targets can drift considerably on the image plane, leading to erroneous tracking outcomes. Addressing such challenges typically requires supplementary appearance cues or Camera Motion Compensation (CMC). While these strategies are effective, they also introduce a considerable computational burden, posing challenges for real-time MOT. In response to this, we introduce UCMCTrack, a novel motion model-based tracker robust to camera movements. Unlike conventional CMC that computes compensation parameters frame-by-frame, UCMCTrack consistently applies the same compensation parameters throughout a video sequence. It employs a Kalman filter on the ground plane and introduces the Mapped Mahalanobis Distance (MMD) as an alternative to the traditional Intersection over Union (IoU) distance measure. By leveraging projected probability distributions on the ground plane, our approach efficiently captures motion patterns and adeptly manages uncertainties introduced by homography projections. Remarkably, UCMCTrack, relying solely on motion cues, achieves state-of-the-art performance across a variety of challenging datasets, including MOT17, MOT20, DanceTrack and KITTI. More details and code are available at https://github.com/corfyi/UCMCTrack

UCMCTrack: Multi-Object Tracking with Uniform Camera Motion Compensation

TL;DR

This paper addresses multi-object tracking under challenging camera motion by designing UCMCTrack, a pure motion-model tracker that operates on the ground plane. It replaces frame-by-frame camera motion compensation with a single sequence-wide parameter set and adopts a Kalman filter on ground-plane states, using the Mapped Mahalanobis Distance to perform data association with ground-plane uncertainty modeled explicitly. Through correlated ground-plane measurements and a process noise compensation scheme, UCMCTrack achieves state-of-the-art results on MOT17, MOT20, DanceTrack, and KITTI, while maintaining real-time efficiency (> FPS on CPU). The work highlights the benefit of grounding motion modeling in the ground plane and suggests future improvements by integrating traditional appearance cues (IoU, ReID) to further enhance robustness and generalization.

Abstract

Multi-object tracking (MOT) in video sequences remains a challenging task, especially in scenarios with significant camera movements. This is because targets can drift considerably on the image plane, leading to erroneous tracking outcomes. Addressing such challenges typically requires supplementary appearance cues or Camera Motion Compensation (CMC). While these strategies are effective, they also introduce a considerable computational burden, posing challenges for real-time MOT. In response to this, we introduce UCMCTrack, a novel motion model-based tracker robust to camera movements. Unlike conventional CMC that computes compensation parameters frame-by-frame, UCMCTrack consistently applies the same compensation parameters throughout a video sequence. It employs a Kalman filter on the ground plane and introduces the Mapped Mahalanobis Distance (MMD) as an alternative to the traditional Intersection over Union (IoU) distance measure. By leveraging projected probability distributions on the ground plane, our approach efficiently captures motion patterns and adeptly manages uncertainties introduced by homography projections. Remarkably, UCMCTrack, relying solely on motion cues, achieves state-of-the-art performance across a variety of challenging datasets, including MOT17, MOT20, DanceTrack and KITTI. More details and code are available at https://github.com/corfyi/UCMCTrack
Paper Structure (33 sections, 26 equations, 4 figures, 5 tables, 1 algorithm)

This paper contains 33 sections, 26 equations, 4 figures, 5 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Distance Measures
  4. Motion Models
  5. Camera Motion Compensation
  6. Method
  7. Motion Modeling on Ground Plane
  8. Correlated Measurement Distribution
  9. Mapped Mahalanobis Distance
  10. Process Noise Compensation
  11. Experiments
  12. Setting
  13. Datasets
  14. Metrics
  15. Implementation Details
  16. ...and 18 more sections

Figures (4)

  • Figure 1: IDF1-HOTA-AssA comparisons of different trackers on the test set of MOT17. The horizontal axis is IDF1,the vertical axis is HOTA, and the radius of circle is AssA. Our UCMCTrack+ achieves 65.8 HOTA, 81.1 IDF1 on MOT17 test, possessing significant competitiveness compared to SOTA trackers. Details are given in Table \ref{['table1']}.
  • Figure 2: The pipeline of the proposed UCMCTrack.
  • Figure 3: Visualization of distance measures. (a) Visualization of IoU on the image plane. IoU fails as there is no intersection between bounding boxes. (b) Visualization of Mapped Mahalanobis Distance (MMD) without Correlated Measurement Distribution (CMD). Incorrect associations occur due to insufficient utilization of distribution information. (c) Visualization of MMD with CMD. Correct associations after using the correlated probability distribution, undergoing a rotation on the ground plane.
  • Figure 4: In-depth analysis of key parameters and robustness in UCMCTrack.