Large vocabulary continuous speech recognition of Broadcast News - The Philips/RWTH approach

Speech Communication

Volumn 37, Issue 1-2, 2002, Pages 109-131

  Beyerlein, P ^a   Aubert, X ^a   Haeb Umbach, R ^a   Harris, M ^a   Klakow, D ^a   Wendemuth, A ^a   Molau, S ^b   Ney, H ^b   Pitz, Michael ^b   Sixtus, A ^b  

^a PHILIPS RESEARCH LABORATORIES   (Germany)

^b RWTH AACHEN UNIVERSITY   (Germany)

Author keywords

Automatic segmentation; Broadcast news; Discriminative model combination; Distance language models; Hub 4; Language model look ahead; Log linear interpolation; Phrases; Script verification; Speaker clustering; Time synchronous one pass trigram decoding; Vocal tract normalization

Indexed keywords

AUDIO ACOUSTICS; BIT ERROR RATE; COMPUTATIONAL METHODS; DATA REDUCTION; DECODING; FEATURE EXTRACTION; INTERPOLATION; MATHEMATICAL MODELS; MICROPHONES; RADIO BROADCASTING;

DISCRIMINATIVE MODEL COMBINATION (DMC);

SPEECH RECOGNITION;

EID: 0036567736     PISSN: 01676393     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0167-6393(01)00062-0     Document Type: Article

References (39)

1. Improvements on the pronunciation prefix tree search organization
2. One pass crossword decoding for large vocabularies based on a lexical tree search organization
3. Large vocabulary continuous speech recognition using word graphs
4. Large vocabulary continuous speech recognition of Wall Street Journal corpus
5. Discriminative model combination
6. Modelling and decoding of crossword context dependent phones in the Philips large vocabulary continuous speech recognition system
7. Automatic transcription of English broadcast news
8. The Philips/RWTH system for transcription of Broadcast News
9. Speaker, environment and channel change detection and clustering via the Bayesian information criterion
10. Generalized iterative scaling for log linear models
11. Comparison of parametric representations for Monosyllabic word recognition in continuously spoken sentences
12. A post-processing system to yield reduced word error rates: Recognizer output voting error reduction (ROVER)
13. An investigation of cepstral parameterisations for large vocabulary speech recognition
14. Segment generatation and clustering in the HTK Broadcast News transcription system
15. Perceptual linear predictive (PLP) analysis of speech
16. Automatic speaker clustering
17. Language-model optimization by mapping of corpora
18. Log-linear interpolation of language models
19. Statistical language modeling using a variable context length
20. The 1997 BBN BYBLOS system applied to Broadcast News transcription
21. Speaker normalization using efficient frequency warping procedures
22. On the probabilistic interpretation of neural network classifiers and discriminative training criteria
23. Improvements in beam search for 10,000-word continuous speech recognition
24. A one pass decoder design for large vocabulary recognition
25. On the limitations of Cepstral features in noise
26. Language-model look-ahead for large vocabulary speech recognition
27. A comparison of time conditioned and word conditioned search techniques for large vocabulary speech recognition
28. A word graph algorithm for large vocabulary continuous speech recognition
29. Capturing long-range correlations using log-linear language models
30. Automatic verification of Broadcast News transcriptions
31. Modeling those F-conditions - Or not
32. Automatic segmentation and clustering of Broadcast News audio
33. Improvements in beam search
34. Speaker adaptation in the philips system for large vocabulary continuous speech recognition
35. A study on speaker normalization using vocal tract normalization and speaker adaptive training
36. Broadcast News transcription using HTK