CONF Hain_INTERSPEECH_2010/IDIAP The AMIDA 2009 Meeting Transcription System Hain, Thomas Burget, Lukas Dines, John Garner, Philip N. El Hannani, Asmaa Huijbregts, Marijn Karafiat, Martin Lincoln, Mike Wan, Vincent EXTERNAL https://publications.idiap.ch/attachments/papers/2010/Hain_INTERSPEECH_2010.pdf PUBLIC Proceedings of Interspeech Makuhari, Japan 2010 September 2010 We present the AMIDA 2009 system for participation in the NIST RT’2009 STT evaluations. Systems for close-talking, far field and speaker attributed STT conditions are described. Im- provements to our previous systems are: segmentation and diar- isation; stacked bottle-neck posterior feature extraction; fMPE training of acoustic models; adaptation on complete meetings; improvements to WFST decoding; automatic optimisation of decoders and system graphs. Overall these changes gave a 6- 13% relative reduction in word error rate while at the same time reducing the real-time factor by a factor of five and using con- siderably less data for acoustic model training.