Code for paper: DBPNet: Dual-Branch Parallel Network with Temporal-Frequency Fusion for Auditory Attention Detection
This paper introduces DBPNet, a novel AAD network that integrates temporal and frequency domains to enhance EEG signal decoding.
Qinke Ni, Hongyu Zhang, Shengbing Pei, Chang Zhou, Zhao Lv, Cunhang Fan. DBPNet: Dual-Branch Parallel Network with Temporal-Frequency Fusion for Auditory Attention Detection. In Ijcai 2024.
- Please download the AAD dataset for training.
- The public KUL dataset, DTU dataset and MM-AAD(not yet open) are used in this paper.
- Python3.11.4
pip install -r requirements.txt
- Modify the
args.*variable in model.py to match the dataset - Using model.py to train and test the model
- Using multi_processing.py to train and test the model in parallel
The First Chinese Auditory Attention Decoding Challenge organized by us at ISCSLP 2024 is now open for registration. The baseline code and data are all public. Everyone is welcome to sign up and participate.
Challenge Website： http://www.iscslp2024.com/ChineseAAD
Timeline of the Chinese AAD Challenge:
21 May 2024: Release of the baseline system, Train and Eval data.
20 Jun 2024: Registration deadline, the due date for participants to join the challenge.
4 Jul 2024: Release of the Test data.
6 Jul 2024: Final submission deadline.
8 Jul 2024: Release of the results and rankings.