ASMSP-P6.2
Splitformer: An improved early-exit architecture for automatic speech recognition on edge devices
Maxence Lasbordes, Universite` Paris-Dauphine, Universite` PSL, Telecom SudParis, Institut Polytechnique de Paris, France; Daniele Falavigna, Alessio Brutti, Fondazione Bruno Kessler, Italy
Session:
ASMSP-P6: Speech Recognition Poster
Track:
ASMSP - Acoustic, Speech and Music Signal Processing
Location:
Poster Area C
Presentation Time:
Wed, 10 Sep, 15:30 - 17:10 Italy Time (UTC +2)
Session Chair:
Daniele Falavigna, Fondazione Bruno Kessler
Presentation
Discussion
Resources
No resources available.
Session ASMSP-P6
ASMSP-P6.1: JOINT BEAMFORMING AND SPEAKER-ATTRIBUTED ASR FOR REAL DISTANT-MICROPHONE MEETING TRANSCRIPTION
Can Cui, iFLYTEK, China; Imran Sheikh, vivoka, France; Mostafa Sadeghi, Emmanuel Vincent, Inria, France
ASMSP-P6.2: Splitformer: An improved early-exit architecture for automatic speech recognition on edge devices
Maxence Lasbordes, Universite` Paris-Dauphine, Universite` PSL, Telecom SudParis, Institut Polytechnique de Paris, France; Daniele Falavigna, Alessio Brutti, Fondazione Bruno Kessler, Italy
ASMSP-P6.3: REFINING TRANSCRIPTS WITH TV SUBTITLES BY PROMPT-BASED WEAKLY SUPERVISED TRAINING OF ASR
Xinnian Zhao, Hugo Van Hamme, KU Leuven University, Belgium
ASMSP-P6.4: BRIDGING THE GAP IN CHILDREN'S SPEECH RECOGNITION: ZERO-SPEECH APPROACHES WITH SPEECH MODIFICATIONS AND ASR ARCHITECTURES
Abhijit Sinha, NIT Sikkim, India; Mittul Singh, AMD Silo AI, Finland; Sudarsana Reddy Kadiri, University of Southern California, India; Hemant Kumar Kathania, NIT Sikkim, India; Mikko Kurimo, AALTO University, India
ASMSP-P6.5: END-TO-END JOINT PUNCTUATED AND NORMALIZED ASR WITH A LIMITED AMOUNT OF PUNCTUATED TRAINING DATA
Can Cui, iFLYTEK, China; Imran Sheikh, vivoka, France; Mostafa Sadeghi, Emmanuel Vincent, Inria, France
ASMSP-P6.6: Model-free Speculative Decoding for Transformer-based ASR with Token Map Drafting
TUAN VU HO, Hiroaki Kokubo, Masaaki Yamamoto, Yohei Kawaguchi, Hitachi, Ltd., Japan
ASMSP-P6.7: ENHANCED SELF-SUPERVISED SPEAKER DIARIZATION FRAMEWORK WITH CONFORMER AND HYBRID CLUSTERING
MALA J B, Alex Raj S M, Rajeev Rajan, APJ Abdul Kalam Technological University, Kerala, India, India
ASMSP-P6.8: LANCET: Lightweight Attention-enhanced Network for Robust Speech Emotion Recognition
Yassin TERRAF, Youssef IRAQI, Mohammed vi polytechnic university, Morocco
ASMSP-P6.9: Pretraining and Adaptation Techniques for Electrolaryngeal Speech Recognition
Lester Phillip Violeta, Ding Ma, Wen-Chin Huang, Tomoki Toda, Nagoya University, Japan