(Image from https://mediapipe.page.link/pose_py_colab)

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 mediapipe_pose_world_landmarks.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 mediapipe_pose_world_landmarks.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 mediapipe_pose_world_landmarks.py --video VIDEO_PATH

By adding the --model option, you can specify model type which is selected from "lite", "full", "heavy".
(default is heavy)

$ python3 mediapipe_pose_world_landmarks.py --model heavy

By specifying the --world_landmark option, POSE_WORLD_LANDMARKS plot is drawn.

$ python3 mediapipe_pose_world_landmarks.py --world_landmark

• MediaPipe
• MediaPipe Pose real-world 3D coordinates

TensorFlow Lite

ONNX opset=12

pose_detection.onnx.prototxt
pose_landmark_lite.onnx.prototxt
pose_landmark_full.onnx.prototxt
pose_landmark_heavy.onnx.prototxt