borarak / emotion-recognition-vgg13

A VGG13 classifier on the FER+ emotion dataset

emotion-recognition-vgg13

A VGG13 classifier on the FER+ emotion dataset

Create FER+ dataset

A detailed Colab notebook is available here to help prepare the dataset

Training and inference

To train a model,

python trainer.py

To inferencer using a trained model on a dataset

python inferece.py

Results

1. Confustion matrix on FERTest2013 (8 emotions and weighted CE loss)

   

2. Confustion matrix on FERTest2013 (5 emotions and down-sampling) dataset with weighted CE loss

   

Resources

[1] E. Barsoum, C. Zhang, C. C. Ferrer, and Z. Zhang, “FER+ : Training Deep Networks for Facial Expression Recognition with Crowd-Sourced Label Distribution.”

[2]