A Bayesian Filter for Multi-view 3D Multi-object Tracking with Occlusion Handling
arXiv:2001.04118 · doi:10.1109/TPAMI.2020.3034435
Abstract
This paper proposes an online multi-camera multi-object tracker that only requires monocular detector training, independent of the multi-camera configurations, allowing seamless extension/deletion of cameras without retraining effort. The proposed algorithm has a linear complexity in the total number of detections across the cameras, and hence scales gracefully with the number of cameras. It operates in the 3D world frame, and provides 3D trajectory estimates of the objects. The key innovation is a high fidelity yet tractable 3D occlusion model, amenable to optimal Bayesian multi-view multi-object filtering, which seamlessly integrates, into a single Bayesian recursion, the sub-tasks of track management, state estimation, clutter rejection, and occlusion/misdetection handling. The proposed algorithm is evaluated on the latest WILDTRACKS dataset, and demonstrated to work in very crowded scenes on a new dataset.
arXiv comment18 pages, 11 figures, TPAMI
References in corpus (6)
- YOLOv3: An Incremental Improvement
- High-Speed Tracking with Kernelized Correlation Filters
- MOTChallenge 2015: Towards a Benchmark for Multi-Target Tracking
- Learning Graphical Model Parameters with Approximate Marginal Inference
- End-to-end Learning of Multi-sensor 3D Tracking by Detection
- Multiple Object Tracking in Unknown Backgrounds with Labeled Random Finite Sets
Cited by in corpus (8)
- Occlusion Handling in Generic Object Detection: A Review
- Visual Multi-Object Tracking with Re-Identification and Occlusion Handling using Labeled Random Finite Sets
- Multi-Scan Multi-Sensor Multi-Object State Estimation
- Linear Complexity Gibbs Sampling for Generalized Labeled Multi-Bernoulli Filtering
- An Overview of Multi-Object Estimation via Labeled Random Finite Set
- Robust multi-sensor Generalized Labeled Multi-Bernoulli filter
- Multi-Objective Multi-Agent Planning for Discovering and Tracking Multiple Mobile Objects
- Online Multi-Object Tracking with delta-GLMB Filter based on Occlusion and Identity Switch Handling