background paper implementation

MAE: https://github.com/facebookresearch/mae

deep-SVDD: https://github.com/lukasruff/Deep-SVDD-PyTorch

SVDD: https://github.com/iqiukp/SVDD-Python

ViT: https://github.com/lucidrains/vit-pytorch

• Download pre-trained model of MSCeleb and move the file to ./models

dataset

• Download RAF-DB dataset and extract the raf-basic dir to ./datasets

• Download AffectNet dataset and extract the AffectNet dir to ./datasets

• Download KDEF dataset and extract the KDEF dir to ./datasets

• The dataset for test(syn-KDEF) recon-image only is here. https://drive.google.com/file/d/1woKlRn6PuvfNGYD1941I8SFw5oR0lrEF/view?usp=sharing

• The dataset for test(syn-KDEF) latent vector(.pt) only is here. https://drive.google.com/file/d/1L6mAI1I_y65nPzBWePlPP90MqlTsQ23T/view?usp=share_link

• The pre-trained model is here. https://drive.google.com/file/d/1S_cTYML_HUnRZ6P1CC8gTK9dRT4oEohv/view?usp=share_link

The location of the label is here. "./datasets/EmoLael/label.txt The location of the images is here. "./datasets/Images/ The location of the PT files is here. "./datasets/PT_files/ The location of the pre-trained model is here. "./checkpoints/

run:

CUDA_VISIBLE_DEVICES=0 python main.py

There is a simple test to Latent-OFER model for a emotion inference:

CUDA_VISIBLE_DEVICES=0 python only_test.py

Our experiment of grad cam++ was based on the package grad-cam 1.3.1, which could be pulled by:

pip install grad-cam==1.3.1

Then, run the following code to dump the visual results. (Need to replace several variables manually.)

python run_grad_cam.py