SPE-26.4
A likelihood ratio based domain adaptation method for E2E models
Chhavi Choudhury, Ankur Gandhe, Xiaohan Ding, Ivan Bulyko, Amazon, United States of America
Session:
Speech Recognition: General Topics II
Track:
Speech and Language Processing
Location:
Gather Area C
Presentation Time:
Mon, 9 May, 22:00 - 22:45 China Time (UTC +8)
Mon, 9 May, 14:00 - 14:45 UTC
Mon, 9 May, 14:00 - 14:45 UTC
Session Chair:
Rohit Prabhavalkar, Google
Session SPE-26
SPE-26.1: EXPLOITING CROSS DOMAIN ACOUSTIC-TO-ARTICULATORY INVERTED FEATURES FOR DISORDERED SPEECH RECOGNITION
Shujie Hu, Shansong Liu, Xurong Xie, Mengzhe Geng, Tianzi Wang, Shoukang Hu, Mingyu Cui, Xunying Liu, Helen Meng, The Chinese University of Hong Kong, Hong Kong
SPE-26.2: Conversational Speech Recognition by Learning Conversation-level Characteristics
Kun Wei, Lei Xie, Audio, Speech and Language Processing Group, School of Computer Science, Northwestern Polytechnical University, China; Yike Zhang, Sining Sun, Long Ma, Tencent, xiaowei, China
SPE-26.3: Exploring Machine Speech Chain for Domain Adaptation
Fengpeng Yue, Tom Ko, Yu Zhang, Southern University of Science and Technology, China; Yan Deng, Lei He, Microsoft China, China
SPE-26.4: A likelihood ratio based domain adaptation method for E2E models
Chhavi Choudhury, Ankur Gandhe, Xiaohan Ding, Ivan Bulyko, Amazon, United States of America
SPE-26.5: RETRIEVING SPEAKER INFORMATION FROM PERSONALIZED ACOUSTIC MODELS FOR SPEECH RECOGNITION
Salima Mdhaffar, Jean-François Bonastre, Natalia Tomashenko, Yannick Estève, LIA - University of Avignon, France; Marc Tommasi, Inria - University of Lille, France
SPE-26.6: NON-AUTOREGRESSIVE END-TO-END AUTOMATIC SPEECH RECOGNITION INCORPORATING DOWNSTREAM NATURAL LANGUAGE PROCESSING
Motoi Omachi, Yuya Fujita, Yahoo Japan Corporation, Japan; Shinji Watanabe, Carnegie Mellon University, Japan; Tianzi Wang, Johns Hopkins University, Japan