deepsort-vehicle-tracker | 2D Vehicle Tracker based on works Simple Online

deepsort-vehicle-tracker is a Python library. deepsort-vehicle-tracker has no bugs, it has no vulnerabilities and it has low support. However deepsort-vehicle-tracker build file is not available. You can download it from GitHub.

This is a 2D Vehicle Tracker based on works Simple Online and Realtime Tracking with a Deep Association Metric (Deep SORT) and Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. For DeepSORT, official implementation of authors is used. The official repository contains deprecated modules, so I fixed them. The implementation is using a feature extractor(CNN) trained on pedestrian images. I trained a new Siamese Network using the vehicle dataset shared by user @John1liu and used it as feature extractor for association. Dataset Link. For Faster-RCNN, I used the official torch implementation with a backbone MobileNet V3 trained on COCO Dataset. Tracker only tracks object with classes "car", "truck" and "bus".