SFAMNet: A Scene Flow Attention-based Micro-expression Network

Neurocomputing | Paper | Bibtex |

(Released on April, 2023)

Results

Performance comparison for micro-expression spotting.

Performance comparison for micro-expression recognition.

Performance comparison for micro-expression analysis.

How to run the code

Step 1) Download the processed_data from:

hidden at the moment

The files are structured as follows:

├─annotation
├─pretrained_weights
├─Utils
├─dataloader.py
├─load_data.py
├─main.py
├─network.py
├─prepare_data.py
├─requirements.txt
├─train.py
├─train_utils.py
├─processed_data

├─CASME_cube_recog_rgbd-flow.pkl
└─CASME_cube_spot_rgbd-flow.pkl

Step 2) Installation of packages using pip

pip install -r requirements.txt

Step 3) Network Training and Evaluation

python main.py

  Note for parameter settings

   --train (True/False)
   --emotion (4/7)

Citation

If you find this work useful for your research, please cite

@article{liong2024sfamnet,
  title={SFAMNet: A scene flow attention-based micro-expression network},
  author={Liong, Gen-Bing and Liong, Sze-Teng and Chan, Chee Seng and See, John},
  journal={Neurocomputing},
  volume={566},
  pages={126998},
  year={2024},
  publisher={Elsevier}
}

Feedback

Suggestions and opinions on this work (both positive and negative) are greatly welcomed. Please contact the authors by sending an email to genbing67@gmail.com or cs.chan at um.edu.my.

License and Copyright

The project is open source under BSD-3 license (see the LICENSE file).

©2023 Center of Image and Signal Processing, Faculty of Computer Science and Information Technology, Universiti Malaya.