SLP-L8.6

Dynamically Slimmable Speech Enhancement Network with Metric-Guided Training

Haixin Zhao, Kaixuan Yang, Nilesh Madhu, Ghent University - imec, Belgium
Session:
SLP-L8: Speech Enhancement Using Subspace Modeling and Learning Criteria Oral
Track:
Speech and Language Processing [SL]
Location:
Room 115
Presentation Time:
Wed, 6 May, 15:40 - 16:00
Presentation
Discussion
Resources
No resources available.
Session SLP-L8
SLP-L8.1: WHAT THE STUDENT LEARNS IN KNOWLEDGE DISTILLATION: A SUBSPACE VIEW AND EVIDENCE ON CONVOLUTIONAL RECURRENT NETWORK
Bo Jin, Timin Li, Department of Electronic Engineering, Tsinghua University, China; Guhan Chen, Department of Statistics and Data Science, Tsinghua University, China; Dongmei Li, Department of Electronic Engineering, Tsinghua University, China
SLP-L8.2: EXPLORING RESOLUTION-WISE SHARED ATTENTION IN HYBRID MAMBA-U-NETS FOR IMPROVED CROSS-CORPUS SPEECH ENHANCEMENT
Nikolai Kühne, Jesper Jensen, Jan Østergaard, Zheng-Hua Tan, Aalborg University, Denmark
SLP-L8.3: MAMBAFORMER: STATE-SPACE AUGMENTED SELF-ATTENTION WITH DOWNUP SAMPLING FOR MONAURAL SPEECH ENHANCEMENT
Shengkui Zhao, Tongyi Lab, Alibaba Group, Singapore, Singapore; Haoxu Wang, Tongyi Lab, Alibaba Group, China, China; Zexu Pan, Tongyi Lab, Alibaba Group, Singapore, Singapore; Yiheng Jiang, Biao Tian, Tongyi Lab, Alibaba Group, China, China; Bin Ma, Tongyi Lab, Alibaba Group, Singapore, Singapore; Xiangang Li, Tongyi Lab, Alibaba Group, China, China
SLP-L8.4: Mixture to Beamformed Mixture: Leveraging Beamformed Mixture as Weak-Supervision for Speech Enhancement and Noise-Robust ASR
Zhong-Qiu Wang, Ruizhe Pang, Southern University of Science and Technology, China
SLP-L8.5: Training Dynamics-Aware Multi-Factor Curriculum Learning for Target Speaker Extraction
Yun Liu, Xuechen Liu, National Institute of informatics, Japan; Xiaoxiao Miao, Duke Kunshan University - DKU, Japan; Junichi Yamagishi, National Institute of informatics, Japan
SLP-L8.6: Dynamically Slimmable Speech Enhancement Network with Metric-Guided Training
Haixin Zhao, Kaixuan Yang, Nilesh Madhu, Ghent University - imec, Belgium
Contacts