ECNU-Cross-Innovation-Lab / LGCCT

(DOI: 10.3390/e24071010) LGCCT: A Light Gated and Crossed Complementation Transformer for Multimodal Speech Emotion Recognition
GNU General Public License v3.0
7 stars 0 forks source link

LGCCT

(DOI: 10.3390/e24071010) LGCCT: A Light Gated and Crossed Complementation Transformer for Multimodal Speech Emotion Recognition We will be very glad if you could cite our paper when you fork our paper :)

ResearchGate: https://www.researchgate.net/publication/362176594_LGCCT_A_Light_Gated_and_Crossed_Complementation_Transformer_for_Multimodal_Speech_Emotion_Recognition

Bilibili: https://www.bilibili.com/video/BV1FU4y1v7yY?is_story_h5=false&p=1&share_from=ugc&share_medium=iphone&share_plat=ios&share_session_id=CA3CF9DD-97E6-42E4-A711-4891B4D0D6FD&share_source=WEIXIN_MONMENT&share_tag=s_i&timestamp=1660885099&unique_k=eugW2Yc&share_times=1

Cite: Liu F, Shen S-Y, Fu Z-W, Wang H-Y, Zhou A-M, Qi J-Y. LGCCT: A Light Gated and Crossed Complementation Transformer for Multimodal Speech Emotion Recognition. Entropy. 2022; 24(7):1010. https://doi.org/10.3390/e24071010

Preparation

processed data can be obtained from here.
the folder structure should be as follows

LGCCT
├── data
│   └── **.pkl
├── modules
│ 
├── src
│ 
└── main.py

Config

the parameters can be set in main.py

Run

python ./main.py

Acknowledgement

Some portion of our code is from MUlT