[NeurIPS Workshop 2019] Official code of the paper "Probabilistic 3D Multi-Object Tracking for Autonomous Driving." First Place of the First NuScenes Tracking Challenge in the AI Driving Olympics Workshop of NeurIPS.
I wanted to clean up the code a bit so I and others can build upon this repo for future tracking efforts
Use YAML config files
Use Dependency injection to select which algorithms and covariances to use (see algorithms.py and covariance.py)
Isolate IOU/distance metric calculation and matching algorithms into separate functions in algorithms.py
Move util functions to utils/geometry_utils.py and generic_utils.py
I tested the code with configurations in the readme (config/AB3DMOT.yaml and config/mahalanobis.yaml) and was able to reproduce the results cited in the README.
Thank you very much for the update suggestions! I will keep this pull request open since we have more than one pull request which may conflict with each other.
Hello! Great work on the project.
I wanted to clean up the code a bit so I and others can build upon this repo for future tracking efforts
I tested the code with configurations in the readme (config/AB3DMOT.yaml and config/mahalanobis.yaml) and was able to reproduce the results cited in the README.