Seq_nms_YOLO

This project is adapted from https://github.com/melodiepupu/seq_nms_yolo to simplify the execution of Seq-NMS and YOLOv2 approaches for video object detection.

Membres: Yunyun SUN, Yutong YAN, Sixiang XU, Heng ZHANG

Adapted by: Daniel Armengod, Sergio Márquez

---

Introduction

This project combines YOLOv2(reference) and seq-nms(reference) to realise real time video detection. This tutorial is designed for Linux with Anaconda.

Steps

1. Open a terminal or Anaconda Prompt.
2. Clone the repository and open the folder:
   • git clone https://github.com/sergiomcgd/seq_nms.git
   • cd seq_nms

Now we are ready to create an Anaconda environment, to do this you can follow steps 3 and 4 or you can just load the environment from the provided .yml file doing: conda env create -f Yolo_env.yml. In case you choose the last option, you should jump to step 5.

3. Create an environment (manually):

   • conda create --name Yolo_env python=2.7 (make sure that you install python 2.7 as all the code is implemented using this Python version.)
   • source activate Yolo_env

4. Install the following libraries:

   • cv2: conda install -c conda-forge opencv
   • matplotlib: conda install matplotlib
   • scipy: conda install -c anaconda scipy
   • tensorflow: conda install -c anaconda tensorflow-gpu
   • pillow: conda install -c anaconda pillow
   • tf_object_detection:conda install -c conda-forge tf_object_detection

5. Make the project (make sure that the environment is activated, if not, activate it with source activate Yolo_env):

   • make
   • Set up the development environment by modifying the PATH and LD_LIBRARY_PATH variables:
     • export PATH=/usr/local/cuda-10.1/bin${PATH:+:${PATH}}
     • export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-10.1/lib64
     • export LIBRARY_PATH=$LIBRARY_PATH:/usr/local/cuda-10.1/lib64

6. Download weights and tiny weights

   • wget https://pjreddie.com/media/files/yolo.weights
   • wget https://pjreddie.com/media/files/yolo-tiny.weights

7. Copy a video file to the video folder (/seq_nms/video)

8. In the video folder (/seq_nms/video) run video2img.py and get_pkllist.py:

   • cd video
   • python video2img.py -i video.avi
   • python get_pkllist.py

9. Return to root folder and run python yolo_seqnms.py to generate output images in video/output;

   • cd ..
   • python yolo_seqnms.py

10. To obtain a video from the resulting output images go to video folder and run img2video.py

• cd video
• python img2video.py -i output

The results would be available in the video folder.

Reference

This project copies lots of code from darknet , Seq-NMS and models.

Some instructions were extracted from: https://docs.nvidia.com/cuda/cuda-quick-start-guide/index.html