LMTformer

Code for the paper 'LMTformer: Facial Depression Recognition with Lightweight Multi-Scale Transformer from Videos'

How to Run

You can directly execute the main.py scripts with your own dataset.

To proceed:

1. Change the load and Path on line 30 and 31 of main.py. load is the csv_load folder in root directory, Path is your AVEC datasets.
2. Change the device on line 39 of main.py to your own device
3. We've also passed the parameters of our trained model for you if you only want to test out model. You can use code:model.load_state_dict(torch.load('best.pt',map_location='cuda:0')) to load our parameters for AVEC2013.

Notes

Before running, ensure the videos are preprocessed to extract the required images.

Kindly note that due to authorization constraints, we are unable to share the AVEC datasets here. Therefore, it is necessary for you to independently extract, crop, align and pretreat the facial data.

Citation