hankkkwu / DeepSORT_YOLOv3

Using DeepSort with YOLOv3 to perform multiple objects tracking

Introduction

The purpose of this project is trying to do multiple objects tracking using YOLOv3 with DeepSORT, and make the people tracking more robust. Thanks for these projects, this project now can support YOLOv3.

https://github.com/nwojke/deep_sort

https://github.com/qqwweee/keras-yolo3

https://github.com/Qidian213/deep_sort_yolov3

Some excellent related work

1. https://github.com/xingyizhou/CenterTrack

2. https://github.com/phil-bergmann/tracking_wo_bnw

3. https://github.com/Zhongdao/Towards-Realtime-MOT

4. https://github.com/ifzhang/FairMOT

5. https://github.com/pjl1995/CTracker

Quick Start

1. Download YOLOv3 or tiny_yolov3 weights and cfg from YOLO website, and put it into model_data folder.

2. Put the video into the test_video folder, and modify the video_file name in demo.py.

3. Run YOLO_DEEP_SORT with command :

   python demo.py
   

Training the model

To train the deep association metric model on your datasets you can reference to cosine_metric_learning approach which is provided as a separate repository.

Result Video

1. Result video on MOT16-13

2. Result video on Taiwan street