Research
My research works focus on deep learning and its application to signal processing (speech, image and video), multi-modal learning, self-supervised learning. I am also quite interested in the applications of deep reinforcement learning such as communication network or social network.
|
|
CPTNN: Cross-Parallel Transformer Neural Network for Speech Enhancement in the Time Domain
Kai Wang ,
Bengbeng, He,
Wei-Ping Zhu
In submission to IEEE/ACM Transactions on Audio, Speech, and Language Processing (TASLP)
PDF/
code Will release after acceptance!
Proposed cross-parallel transformer neural network to extract and dynamically fuse the local and global information of long-range speech sequences.
|
|
SE-Mixer: Towards An Efficient Attention-free Neural Network for Speech Enhancement
Kai Wang ,
Bengbeng, He,
Wei-Ping Zhu
In submission to IEEE Signal Processing Letters (SLP)
PDF/
code Will release after acceptance!
Proposed attention-free MLP architecture to achieve competitive performance compared with attention-based methods.
|
|
SE-TransUNet: Transformers Make A Strong UNet for Time Domain Speech Enhancement
Bengbeng, He,
Kai Wang ,
Wei-Ping Zhu
In submission to European Signal Processing Conference (EURASIP), 2022
PDF/
code Will release after acceptance!
Incorporate transformer structure into encoder and decoder of UNet to solve the limited receptive field of the CNN structure
|
|
TSTNN: Two-Stage Transformer Based Neural Network for Speech Enhancement in Time Domain
Kai Wang ,
Bengbeng, He,
Wei-Ping Zhu
IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP), 2021
arXiv/
code
Proposed two-stage transformer neural network to extract local and global information of long-range speech feature.
|
|
CAUNet: Context-Aware UNet for Speech Enhancement in Time Domain
Kai Wang ,
Bengbeng, He,
Wei-Ping Zhu
IEEE International Conference on Symposium on Circuits and Systems (ISCAS), 2021
PDF/
code
Proposed context-aware UNet to extract contextual information of speech feature.
|
Service
Conference reviewer: ISCAS 2021, NEWCAS 2021, NEWCAS 2022
|
This website is inspired by Jon Barron. Thanks!
(last update: Feb 2022)
|
|