Efficient codebooks for fast and accurate low resource ASR systems

Speech Communication

Volumn 51, Issue 9, 2009, Pages 732-743

  Zouari, Leila ^a   Chollet, Gérard ^a  

^a INSTITUT TELECOM   (France)

Author keywords

Codebook; Gaussian selection; Speech recognition

Indexed keywords

AUTOMATIC SPEECH RECOGNITION; CODEBOOK; CODEBOOKS; CONTINUOUS HIDDEN MARKOV MODEL; GAUSSIAN SELECTION; LARGE VOCABULARY CONTINUOUS SPEECH RECOGNITION; LIKELIHOOD COMPUTATION; MULTI-LEVEL; RECOGNITION SPEED; RESOURCE CONSUMPTION; SPEECH INTERFACE; TIME-CONSUMING PARTS;

CONTINUOUS SPEECH RECOGNITION; HIDDEN MARKOV MODELS; MOBILE DEVICES; REMELTING; SPEECH ANALYSIS;

CONTINUOUS TIME SYSTEMS;

EID: 67649134955     PISSN: 01676393     EISSN: None     Source Type: Journal    
DOI: 10.1016/j.specom.2009.01.010     Document Type: Article

References (32)

1. Aiyer A., Gales M.J.F., and Picheny M.A. Rapid likelihood calculation of subspace clustered Gaussian components. Internat. Conf. Acoust. Speech Signal Process. (2000) 1519-1522
2. Bocchieri E. Vector quantization for the efficient computation of continuous density likelihoods. Internat. Conf. Acoust. Speech Signal Process. (1993) 692-695
3. Chan, A., Sherwani, J., Mosur, R., Rudnicky, A., 2004. Four layer categorization scheme of fast GMM computation techniques in large vocabulary continuous speech recognition systems. In: Internat. Conf. on Spoken Language Processing, Jesu Island, Korea.
4. Chan, A., Ravishankar, M., Rudnicky, A., 2005. On improvements to CI based GMM selection. In: Eur. Conf. on Speech Communication and Technology, Lisbon, September 2005, pp. 565-568.
5. Digalakis V., Monaco P., and Murveit H. Genones: generalized mixture tying in continuous Hidden Markov model-based speech recognizers. IEEE Trans. Speech Audio Process. 4 (1996) 281-289
6. Digalakis V., Tsakalidis S., Harizakis c., and Neumeyer L. Efficient speech recognition using subvector quantization and discrete-mixture HMMs. Comput. Speech Lang. (2000) 33-46
7. Filali, K., Li, X., Bilmes, J., 2002. Data-driven vector clustering for low memory footprint ASR. In: Internat. Conf. on Spoken Language Processing.
8. Gales, M.J.F., McKnill, K., Young, S., 1996. Use of Gaussian selection in large vocabulary continuous speech recognition using HMMs. In: Internat. Conf. on Spoken Language Processing, pp. 470-473.
9. Gales M.J.F., McKnill K., and Young S. State based Gaussian selection in large vocabulary continuous speech recognition using HMMs. IEEE Trans. Speech Audio Process. (1999) 470-473
10. Galliano, S., Geoffrois, E., Mostefa, D., Choukri, K., Bonastre, J.F., Gravier, G., 2005. The Ester Phase II campaign for the rich transcription of French broadcast news. In: Eur. Conf. on Speech Communication and Technology.
11. Herman S.H., and Sukkar R.A. Joint MCE estimation of VQ and HMM parameters for Gaussian Mixture Selection. IEEE Internat. Conf. Acoust. Speech Signal Process. (1998) 485-488
12. Jurgen F., and Ivica R. The bucket box intersection (BBI) algorithm for fast approximative evaluation of diagonal mixture Gaussians. Internat. Conf. Acoust. Speech Signal Process. (1996) 837-840
13. Jurgen, F., Ivica, R., Tile, S., 1996. Speeding up the score computation of HMM speech recognizers with the Bucket Voronoi intersection algorithm. In: Eur. Conf. on Speech Communication and Technology, pp. 1091-1094.
14. Lee A. Gaussian mixture selection using context Independent HMM. IEEE Internat. Conf. Acoust. Speech Signal Process (1997)
15. Lee A., Kawahara T., Takeda K., and Shikano K. A new phonetic tied-mixture model for efficient decoding. Internat. Conf. Acoust. Speech Signal Process (2001) 1269-1272
16. Leppänen J., and Kiss I. Gaussian selection with non-overlapping clusters for ASR in embedded devices. Internat. Conf. Acoust. Speech Signal Process (2006)
17. Li X., Malkin J., and Bilmes J. A high-speed, low-resource ASR back-end based on custom arithmetic. IEEE Trans. Audio Speech Language Process. (2006)
18. Mak B., and Bocchieri E. Subspace distribution clustering Hidden Markov model. IEEE Trans. Speech Audio Process. (2001) 264-275
19. Mokbel C. Online adaptation of HMMs to real life conditions: a unified framework. IEEE Trans. Speech Audio Process. (2001) 342-357
20. Mosur, R.M., Bisiani, R., Thayer, E., 1997. Sub-vector clustering to improve memory and speed performance of acoustic likelihood computation. In: Eur. Conf. on Speech Communication and Technology, Rhodes, Greece.
21. Olsen, J., 2000. Gaussian selection using multiple quantisation indexes. In: IEEE Nordic Processing Symposium.
22. Ortmanns, S., Ney, H., Firslaff, T., 1997. Fast likelihood computation methods for continuous mixture densities in large vocabulary speech recognition. In: Eur. Conf. on Speech Communication and Technology, Rhodes, Greece, pp. 139-142.
23. Padmanablan, M., Jan, E., Bahl, L., Picheny, M., 1997. Decision-tree based feature space quantization for fast Gaussian computation. In: IEEE Workshop on Automatic Speech Recognition and Understanding, 1997, Santa Barbara, pp. 325-330.
24. Padmanablan M., Bahl L., and Nahamoo D. Partitioning the feature space of a classifier with linear hyperplanes. IEEE Trans. Speech Audio Process. (1999) 282-288
25. Pellom B., Sarikaya R., and Hansen J.H.L. Fast likelihood computation techniques for nearest-neighbor based search for continuous speech recognition. IEEE Signal Process. Lett. 8 8 (2001) 221-224
26. Sagayama S., and Takahashi S. On the use of scalar quantization for fast HMM computation. Internat. Conf. Acoust. Speech Signal Process. 2 (1995) 213-216
27. Sankar, A., Ramana, V., 1999. Parameter tying and Gaussian clustering for faster, better, and smaller speech recognition. In: Eur. Conf. on Speech, Communication and Technology, Greece.
28. Sankar A., Ramana V., Stolcke A., and Weng F. Improved modeling and efficiency for automatic transcription of broadcast news. Speech Comm. 37 (2002) 133-158
29. Suontausts, J., Hakkinen, J., Viikki, O., 1999. Fast decoding techniques for practical realtime speech recognition systems. In: Workshop on Automatic Speech Recognition and Understanding.
30. Takahashi S., and Sagayama S. Four-level tied-structure for efficient representation of acoustic modeling. Internat. Conf. Acoust. Speech Signal Process. 1 (1995) 520-523
31. Tsakalidis S., Digalakis V., and Neumeyer L. Efficient speech recognition using subvector quantization and discrete-mixture HMMs. IEEE Internat. Conf. Acoust. Speech Signal Process. (1999) 569-572
32. Woszczyna, M., 1998. Fast speaker independent large vocabulary speech recognition. Ph.D. Thesis, Karlsruhe University.