| Paper: | SP-P11.9 | ||
| Session: | Topics in Large Vocabulary Continuous Speech Recognition | ||
| Time: | Thursday, May 20, 09:30 - 11:30 | ||
| Presentation: | Poster | ||
| Topic: | Speech Processing: Large Vocabulary Recognition/Search | ||
| Title: | ADVANCES IN THE AUTOMATIC TRANSCRIPTION OF LECTURES | ||
| Authors: | Mauro Cettolo; ITC-irst | ||
| Fabio Brugnara; ITC-irst | |||
| Marcello Federico; ITC-irst | |||
| Abstract: | Transcribing lectures is a challenging task, both in acoustic and in language modeling. In this work, we present recent results on the automatic transcription of lectures from the TED corpus, released by ELRA and LDC in 2002, which contains the recordings of talks given in English at Eurospeech '93, by mostly non-native speakers.Concerning acoustic modeling, the set of AMs trained for a broadcast news transcription task was adapted on the TED training data through MLLR adaptation, including models of spontaneous speech phenomena. Moreover, a normalization procedure was embodied in the training stage, consisting in a cluster-based mean and variance normalization of the static features.On the side of language modeling, the most effective adaptation of the background language model, estimated on broadcast news transcripts, conference proceedings, lecture transcripts, and conversational speech transcripts, was obtained by exploiting the paper presented in each lecture to be processed.The best transcription performance on a 2 hours test set was 32.4% WER. | ||
| Back | |||
Home -||-
Organizing Committee -||-
Technical Committee -||-
Technical Program -||-
Plenaries
Paper Submission -||-
Special Sessions -||-
ITT -||-
Paper Review -||-
Exhibits -||-
Tutorials
Information -||-
Registration -||-
Travel Insurance -||-
Housing -||-
Workshops