BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Automatic Speech Recognition in a State-of-Flux - Ralf Schlüter\,
  RWTH Aachen
DTSTART:20210622T110000Z
DTEND:20210622T120000Z
UID:TALK160567@talks.cam.ac.uk
CONTACT:Dr Kate Knill
DESCRIPTION:*Abstract*: Initiated by the successful utilization of deep ne
 ural network modeling for large vocabulary automatic speech recognition (A
 SR)\, the last decade brought a considerable diversification of ASR archit
 ectures. Following the classical state-of-the-art hidden Markov model (HMM
 ) based architecture\, connectionist temporal classification (CTC)\, atten
 tion-based encoder-decoder\, recurrent neural network transducer (RNN-T) a
 nd monotonic variants\, as well as segmental approaches including inverted
  HMM architectures were introduced. All these architectures show competiti
 ve performance and the question arises\, which of these will finally preva
 il and define the new state-of-the-art in large vocabulary ASR? In this pr
 esentation\, a comparative review of current architectures in the context 
 of Bayes decision rule is provided. Relations and equivalences between arc
 hitectures are derived\, utilization of data is considered and especially 
 the role of language modeling within integrated end-to-end architectures w
 ill be discussed.\n\n*Bio*: Ralf Schlüter serves as Academic Director and
  Lecturer (Privatdozent) in the Department of Computer Science of the Facu
 lty of Computer Science\, Mathematics and Natural Sciences at RWTH Aachen 
 University. He leads the Automatic Speech Recognition Group at the Lehrstu
 hl Informatik 6: Human Language Technology and Pattern Recognition. He stu
 died physics at RWTH Aachen University and Edinburgh University and receiv
 ed his Diploma in Physics (1995)\,  in Computer Science (2000) and Habilit
 ation for Computer Science (2019)\, each at RWTH Aachen University. Dr. Sc
 hlüter works on all aspects of automatic speech recognition and has been 
 leading the scientific work of the Lehrstuhl Informatik 6 in the area of a
 utomatic speech recognition in many large national and international resea
 rch projects\, e.g. EU-Bridge and TC-STAR (EU)\, Babel (US-IARPA) and Quae
 ro (French OSEO).\n\nThis talk is provided through the "ISCA International
  Virtual Seminar Programme":https://www.isca-speech.org/iscaweb/index.php/
 distinguished-lecturers/online-seminars.\n
LOCATION:Zoom: https://zoom.us/j/95352633552?pwd=RzJVK2UzOGZyNU5mVHd1Y1VPT
 2tDUT09
END:VEVENT
END:VCALENDAR