Fusion of Head and Full-Body Detectors for Multi-Object Tracking

Abstract

In order to track all persons in a scene, the tracking-by-detection paradigm has proven to be a very effective approach. Yet, relying solely on a single detector is also a major limitation, as useful image information might be ignored. Consequently, this work demonstrates how to fuse two detectors into a tracking system. To obtain the trajectories, we propose to formulate tracking as a weighted graph labeling problem, resulting in a binary quadratic program. As such problems are NP-hard, the solution can only be approximated. Based on the Frank-Wolfe algorithm, we present a new solver that is crucial to handle such difficult problems. Evaluation on pedestrian tracking is provided for multiple scenarios, showing superior results over single detector tracking and standard QP-solvers. Finally, our tracker ranks 2nd on the MOT16 benchmark and 1st on the new MOT17 benchmark, outperforming over 90 trackers.

arXiv comment10 pages, 4 figures; Winner of the MOT17 challenge; CVPRW 2018

References in corpus (9)

• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
• MOT16: A Benchmark for Multi-Object Tracking
• Tracking the Trackers: An Analysis of the State of the Art in Multiple Object Tracking
• Convergence Rate of Frank-Wolfe for Non-Convex Objectives
• A Multi-cut Formulation for Joint Segmentation and Tracking of Multiple Objects
• Binary Quadratic Programing for Online Tracking of Hundreds of People in Extremely Crowded Scenes
• Tracking with multi-level features
• Learning to Divide and Conquer for Online Multi-Target Tracking
• Re-identification of Humans in Crowds using Personal, Social and Environmental Constraints

Cited by in corpus (10)

• Deep Continuous Conditional Random Fields with Asymmetric Inter-object Constraints for Online Multi-object Tracking
• Tracklet Association Tracker: An End-to-End Learning-based Association Approach for Multi-Object Tracking
• How To Train Your Deep Multi-Object Tracker
• Exploit the Connectivity: Multi-Object Tracking with TrackletNet
• Multi-Object Tracking with Siamese Track-RCNN
• Multiple People Tracking Using Hierarchical Deep Tracklet Re-identification
• MOTR: End-to-End Multiple-Object Tracking with Transformer
• Learning a Proposal Classifier for Multiple Object Tracking
• Multiple Object Tracking with Mixture Density Networks for Trajectory Estimation
• SiamMOT: Siamese Multi-Object Tracking