anime-face-detector

Anime Face Detector

Input

(Image from https://github.com/hysts/anime-face-detector/blob/main/assets/input.jpg)

• face_yolov3 input shape : (1, 3, 608, 608)
• face_faster-rcnn input shape : (1, 3, 800, 1333)
• face_hrnetv2 input shape : (n, 3, 256, 256)

Output

• dets shape : (1, 100, 5)
• heatmap shape : (n, 28, 64, 64)

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 anime-face-detector.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 anime-face-detector.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 anime-face-detector.py --video VIDEO_PATH

By adding the --detector option, you can specify face detector model which is selected from "yolov3", "faster-rcnn". (default is yolov3)

$ python3 anime-face-detector.py --detector yolov3

Reference

• Anime Face Detector
• MMCV
• MMDetection
• MMPose

Framework

Pytorch

Model Format

ONNX opset=11

Netron

anime-face_yolov3.onnx.prototxt
anime-face_faster-rcnn.onnx.prototxt
anime-face_hrnetv2.onnx.prototxt