openpose

OpenPose

Input

(Image from https://pixabay.com/ja/photos/%E5%A5%B3%E3%81%AE%E5%AD%90-%E7%BE%8E%E3%81%97%E3%81%84-%E8%8B%A5%E3%81%84-%E3%83%9B%E3%83%AF%E3%82%A4%E3%83%88-5204299/)

Input shape : (1, 3, 240, 320)
Range : [0, 255]

Output

• Confidence : (1, 19, 30, 40)
• Range : [0, 1.0]

Note

OPENPOSE: MULTIPERSON KEYPOINT DETECTION SOFTWARE LICENSE AGREEMENT ACADEMIC OR NON-PROFIT ORGANIZATION NONCOMMERCIAL RESEARCH USE ONLY

Usage

Automatically downloads the onnx and prototxt files on the first run. It is necessary to be connected to the Internet while downloading.

For the sample image,

$ python3 openpose.py

If you want to specify the input image, put the image path after the --input option.
You can use --savepath option to change the name of the output file to save.

$ python3 openpose.py --input IMAGE_PATH --savepath SAVE_IMAGE_PATH

By adding the --video option, you can input the video. If you pass 0 as an argument to VIDEO_PATH, you can use the webcam input instead of the video file.

$ python3 openpose.py --video VIDEO_PATH

Also you can set detection threshold by adding the --threshold option. Default value is 0.3.

$ python3 openpose.py --threshold 0.2

By adding the --single_scale option, you can use single scale detection for performance.

$ python3 openpose.py --single_scale

Reference

Code repo for realtime multi-person pose estimation in CVPR'17 (Oral)

Framework

Caffe

Model Format

CaffeModel

Netron

pose_deploy.prototxt