International Workshop on Spoken Language Translation (IWSLT) 2012

Hong Kong
December 6-7, 2012

The 2012 KIT and KIT-NAIST English ASR Systems for the IWSLT Evaluation

Christian Saam (1), Christian Mohr (1), Kevin Kilgour (1), Michael Heck (1), Matthias Sperber (1), Keigo Kubo (2), Sebastian Stüker (1), Sakriani Sakti (2), Graham Neubig (2), Tomoki Toda (2), Satoshi Nakamura (2), Alex Waibel (1)

(1) Institute for Anthropomatics, Karlsruhe Institute of Technology, Germany
(2) Augmented Human Communication Laboratory, Nara Institute of Science and Technology, Japan

This paper describes our English Speech-to-Text (STT) systems for the 2012 IWSLT TED ASR track evaluation. The systems consist of 10 subsystems that are combinations of different front-ends, e.g. MVDR based and MFCC based ones, and two different phone sets. The outputs of the subsystems are combined via confusion network combination. Decoding is done in two stages, where the systems of the second stage are adapted in an unsupervised manner on the combination of the first stage outputs using VTLN, MLLR, and cMLLR. Index Terms: speech recognition, IWSLT, TED talks, evaluation system, system development

Full Paper   

Bibliographic reference.  Saam, Christian / Mohr, Christian / Kilgour, Kevin / Heck, Michael / Sperber, Matthias / Kubo, Keigo / Stüker, Sebastian / Sakti, Sakriani / Neubig, Graham / Toda, Tomoki / Nakamura, Satoshi / Waibel, Alex (2012): "The 2012 KIT and KIT-NAIST English ASR systems for the IWSLT evaluation", In IWSLT-2012, 87-90.