CS-766---Computer-Vision-Project

Object detection and tracking for automated vehicles This is the project of course CS766 - Computer Vision. We implement yolov5 + Deepsort and EfficientDet + Deepsort for object detection and tracking for automated vehicles. The work can be summarized based on the following steps:

1. Re-implement the YOLO V5 algorithm + DeepSort algorithm and EfficientDet algorithm + Deepsort algorithm. The pre-trained weights are based on the COCO dataset.
2. Use transferLearning to develop models based on the new self-driving datasets
3. Evaluate the YOLOv5 model for object detection. Compare the pre-trained models and models after transferLearning.
4. Evaluate DeepSort Algorithm based on YOLOv5 and EfficientDet.

The project is inspired by: https://github.com/mikel-brostrom/Yolov5_DeepSort_OSNet; https://github.com/ImLaoBJie/EfficientDet_Deepsort.