Skip to content

Latest commit

 

History

History

facemesh_v2

README.md

FaceMesh-V2

Input

(Image from https://storage.googleapis.com/mediapipe-assets/portrait.jpg)

FaceDetector

  • ailia input shape: (1, 128, 128, 3) RGB channel order
  • Pixel value range: [-1, 1]

FaceMesh-V2

  • ailia input shape: (1, 256, 256, 3) RGB channel order
  • Pixel value range: [0, 1]

Blendshape

  • ailia input shape: (1, 146, 2)

Output

Face landmark

Blendshape

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 facemesh_v2.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 facemesh_v2.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 facemesh_v2.py --video VIDEO_PATH --savepath SAVE_VIDEO_PATH

If you specify the --blendshape option, a bar plot of Blendshape will be output.

$ python3 facemesh_v2.py --blendshape

Reference

Framework

TensorFlow Lite

Model Format

ONNX opset = 11

Netron