2021 IEEE International Conference on Acoustics, Speech and Signal Processing

6-11 June 2021 • Toronto, Ontario, Canada

Extracting Knowledge from Information

2021 IEEE International Conference on Acoustics, Speech and Signal Processing

6-11 June 2021 • Toronto, Ontario, Canada

Extracting Knowledge from Information

Technical Program

SPE-39: Speech Recognition 13: Acoustic Modeling 1

Session Type: Poster
Time: Thursday, 10 June, 15:30 - 16:15
Location: Gather.Town
Session Chair: Xiaodong Cui, IBM
 
SPE-39.1: SPEECH ACOUSTIC MODELLING FROM RAW PHASE SPECTRUM
         Erfan Loweimi; University of Edinburgh
         Zoran Cvetkovic; King's College London
         Peter Bell; University of Edinburgh
         Steve Renals; University of Edinburgh
 
SPE-39.2: AN INVESTIGATION OF USING HYBRID MODELING UNITS FOR IMPROVING END-TO-END SPEECH RECOGNITION SYSTEM
         Shunfei Chen; Hithink RoyalFlush AI Research Institute
         Xinhui Hu; Hithink RoyalFlush AI Research Institute
         Sheng Li; National Institute of Information and Communications Technology
         Xinkang Xu; Hithink RoyalFlush AI Research Institute
 
SPE-39.3: FEDERATED ACOUSTIC MODELING FOR AUTOMATIC SPEECH RECOGNITION
         Xiaodong Cui; IBM T. J. Watson Research Center
         Songtao Lu; IBM T. J. Watson Research Center
         Brian Kingsbury; IBM T. J. Watson Research Center
 
SPE-39.4: EAT: ENHANCED ASR-TTS FOR SELF-SUPERVISED SPEECH RECOGNITION
         Murali Karthick Baskar; Brno University of Technology
         Lukáš Burget; Brno University of Technology
         Shinji Watanabe; Johns Hopkins University
         Ramon Astudillo; IBM T. J. Watson Research Center
         Jan "Honza" Cernocky; Brno University of Technology
 
SPE-39.5: NEURAL ARCHITECTURE SEARCH FOR LF-MMI TRAINED TIME DELAY NEURAL NETWORKS
         Shoukang Hu; The Chinese University of Hong Kong
         Xurong Xie; The Chinese University of Hong Kong
         Shansong Liu; The Chinese University of Hong Kong
         Mingyu Cui; The Chinese University of Hong Kong
         Mengzhe Geng; The Chinese University of Hong Kong
         Xunying Liu; The Chinese University of Hong Kong
         Helen Meng; The Chinese University of Hong Kong
 
SPE-39.6: HYPOTHESIS STITCHER FOR END-TO-END SPEAKER-ATTRIBUTED ASR ON LONG-FORM MULTI-TALKER RECORDINGS
         Xuankai Chang; Johns Hopkins University
         Naoyuki Kanda; Microsoft Corporation
         Yashesh Gaur; Microsoft Corporation
         Xiaofei Wang; Microsoft Corporation
         Zhong Meng; Microsoft Corporation
         Takuya Yoshioka; Microsoft Corporation