Event
GenTrack2 Multi-Object Tracking Method Paper Released
1 reports · 1 sources
Toan Van Nguyen and other authors released a paper proposing a visual multi-object tracking method, GenTrack2, which combines stochastic particle filtering with deterministic association, with the goal of maintaining identity consistency under conditions where the number of targets is unknown and changes over time and the dynamics are nonlinear. Methodologically, the particle filter uses particle swarm optimization to guide particles to converge toward the modes of the state distribution, with fitness integrating motion consistency, appearance similarity, and social interaction cues with neighboring targets; the deterministic association strengthens identity consistency through a cost matrix that includes the spatial consistency between particles and current detections, detection confidence, and trajectory penalties. The authors state that the method can be used both for pre-recorded video and for real-time camera streams where future frames are unavailable, and that reference implementations of the method and comparison trackers have been made public on GitHub.
Report timeline
- GenTrack2 proposes a multi-object tracking method combining stochastic and deterministic approaches
GenTrack2 proposes a visual multi-object tracking method that jointly uses stochastic particle filtering and deterministic association to maintain identifier consistency under conditions where the number of targets is unknown and changes over time and the dynamics are nonlinear. The particle filter uses particle swarm optimization (PSO) to guide particles toward the modes of the state distribution, with a fitness measure that combines motion consistency, appearance similarity, and social-interaction cues with neighboring targets; deterministic association strengthens identifier consistency through a cost matrix that incorporates spatial consistency between particles and current detections, detection confidence, and track penalties.
arXiv cs.RO ·