/pose_estimation

Primary LanguagePython

Multi Agent Pose Estimation

The code takes a video and adds pose information onto it. For best usage ensure that there aren't many people in the frame.

Requirements

  1. opencv (pip install opencv-contrib-python)
  2. ffmpeg

Usage

Place the video in question in the same folder and then:

  1. Adjust the input name, length of video required in run.sh. Defaults to dancing.mp4 and 30 seconds otherwise.
  2. Run bash run.sh.

References

  1. Adapted from: https://www.learnopencv.com/multi-person-pose-estimation-in-opencv-using-openpose