SPE-30.5
INVESTIGATION OF ROBUSTNESS OF HUBERT FEATURES FROM DIFFERENT LAYERS TO DOMAIN, ACCENT AND LANGUAGE VARIATIONS
Pratik Kumar, Vrunda N. Sukhadia, Srinivasan Umesh, Indian Institute of Technology Madras, India
Session:
Language Identification and Low Resource Speech Recognition
Track:
Speech and Language Processing
Location:
Gather Area C
Presentation Time:
Mon, 9 May, 23:00 - 23:45 China Time (UTC +8)
Mon, 9 May, 15:00 - 15:45 UTC
Mon, 9 May, 15:00 - 15:45 UTC
Session Co-Chairs:
Abdelrahman Mohamed, Meta and Jochen Ehnes, Institute for Infocomm Research, A*STAR
Session SPE-30
SPE-30.1: Spoken language recognition with cluster-based modeling
Stanisław Kacprzak, Magdalena Rybicka, Konrad Kowalczyk, Akademia Górniczo-Hutnicza im. Stanisława Staszica w Krakowie, Poland
SPE-30.2: PHONOTACTIC LANGUAGE RECOGNITION USING A UNIVERSAL PHONEME RECOGNIZER AND A TRANSFORMER ARCHITECTURE
David Romero, Christian Salamea, Universidad Politecnica Salesiana, Ecuador; Luis Fernando D'Haro, Marcos Estecha-Garitagoitia, Universidad Politécnica de Madrid, Spain
SPE-30.3: IMPROVED LANGUAGE IDENTIFICATION THROUGH CROSS-LINGUAL SELF-SUPERVISED LEARNING
Andros Tjandra, Diptanu Gon Choudhury, Frank Zhang, Kritika Singh, Alexis Conneau, Alexei Baevski, Assaf Sela, Yatharth Saraf, Michael Auli, Facebook AI, United States of America
SPE-30.4: LANGUAGE ADAPTIVE CROSS-LINGUAL SPEECH REPRESENTATION LEARNING WITH SPARSE SHARING SUB-NETWORKS
Yizhou Lu, Mingkun Huang, Xinghua Qu, Pengfei Wei, Zejun Ma, ByteDance AI Lab, China
SPE-30.5: INVESTIGATION OF ROBUSTNESS OF HUBERT FEATURES FROM DIFFERENT LAYERS TO DOMAIN, ACCENT AND LANGUAGE VARIATIONS
Pratik Kumar, Vrunda N. Sukhadia, Srinivasan Umesh, Indian Institute of Technology Madras, India
SPE-30.6: COMBINING UNSUPERVISED AND TEXT AUGMENTED SEMI-SUPERVISED LEARNING FOR LOW RESOURCED AUTOREGRESSIVE SPEECH RECOGNITION
Chak-Fai Li, Francis Keith, William Hartmann, Matthew Snover, Raytheon BBN Technologies, United States of America