hsemotion

High-Speed face Emotion recognition

Input

• Ailia input shape: (1, 3, 224, 224) RGB channel order
  • Ailia input shape: (1, 3, 260, 260) for B2 model
• Pixel value range: [0, 1] before normalization
• Preprocessing: normalization using ImageNet statistics

Output

emotion_class_count=4
+ idx=0
  category=5 [ Neutral ]
  prob=0.6248039603233337
+ idx=1
  category=4 [ Happiness ]
  prob=0.15010859072208405
+ idx=2
  category=7 [ Surprise ]
  prob=0.07648341357707977
+ idx=3
  category=6 [ Sadness ]
  prob=0.05946649610996246

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 hsemotion.py 

If you want to specify the input image, put the image path after the --input option.

$ python3 hsemotion.py --input 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 hsemotion.py --video VIDEO_PATH

Reference

High-Speed face Emotion recognition

Framework

PyTorch

Model Format

ONNX opset = 11

Netron

• enet_b0_8_best_afew.onnx.prototxt
• enet_b0_8_best_vgaf.onnx.prototxt
• enet_b0_8_va_mtl.onnx.prototxt
• enet_b2_8.onnx.prototxt