EAVFormer

Our article is under review and we will open source all the code in the future.

Environment configuration

The code is trained andtested with Python=3.10.0, pytorch=1.13.0, torchaudio=0.13.0, opencv-python=4.8.0, scikit-learn=1.3.0, timm=0.9.5, matplotlib=3.7.2.

Weight file

Links：https://pan.baidu.com/s/1E10aKKbFudwO4EEMkbkQpw?pwd=uqwc Extraction code：uqwc

Dataset Download

Download the dataset via the following web link or web disk link.

• RAVDESS：https://zenodo.org/records/1188976
• CMU-MOSEI：http://immortal.multicomp.cs.cmu.edu/
• CREMA-D: https://github.com/CheyneyComputerScience/CREMA-D