Out-of-the-box code and models for CMU's object detection and tracking system for multi-camera surveillance videos. Speed optimized Faster-RCNN model. Tensorflow based. Also supports EfficientDet. WACVW'20
See the --tracking_objs parameter here. The default is Person,Vehicle. You can simply add additional object class names (comma separated) according to the name list here. You can only do this with VIRAT trained object models (v1-6).
This model is tracking "Person" and "Vehicle" class. How to add Bicycle etc. clase? Also if I want to add other classes, which code should I change?