Investigation on dimensionality reduction of concatenated features with deep neural network for LVCSR systems

International Conference on Signal Processing Proceedings, ICSP

Volumn 1, Issue , 2012, Pages 562-566

  Bao, Yebo ^a   Jiang, Hui ^b   Liu, Cong ^c   Hu, Yu ^c   Dai, Lirong ^a  

^a UNIVERSITY OF SCIENCE AND TECHNOLOGY OF CHINA   (China)

^b YORK UNIVERSITY   (Canada)

^c IFLYTEK RESEARCH   (China)

Author keywords

Bottleneck features; Deep neural networks; Dimensionality reduction; Large vocabulary continuous speech recognition

Indexed keywords

BOTTLENECK FEATURES; DEEP NEURAL NETWORKS; DIMENSIONALITY REDUCTION; DIMENSIONALITY REDUCTION TECHNIQUES; HIGH-DIMENSIONAL FEATURES; LARGE VOCABULARY CONTINUOUS SPEECH RECOGNITION; LOW-DIMENSIONAL REPRESENTATION; RECOGNITION PERFORMANCE;

CONTINUOUS SPEECH RECOGNITION; HIDDEN MARKOV MODELS; SIGNAL PROCESSING; VOCABULARY CONTROL;

NEURAL NETWORKS;

EID: 84876477729     PISSN: None     EISSN: None     Source Type: Conference Proceeding    
DOI: 10.1109/ICoSP.2012.6491550     Document Type: Conference Paper

References (12)

1. A. Mohamed, G. E. Dahl, and G. E. Hinton, "Deep belief networks for phone recognition," in NIPS Workshop on Deep Learning for Speech Recognition and Related Applications, 2009.
2. F. Seide, G. Li, and D. Yu, "Conversational speech transcription using context-dependent deep neural networks," in Interspeech, 2011, pp. 437- 440.
3. G. E. Dahl, D. Yu, L. Deng, and A. Acero, "Context-dependent pretrained deep neural networks for large vocabulary speech recognition," IEEE Trans. on Audio, Speech and Language Processing, January 2012.
4. Jia Pan, Cong Liu, Zhiguo Wang, Yu Hu, Hui Jiang, "Investigation of Deep Neural Networks (DNN) for Large Vocabulary Continuous Speech Recognition: Why DNN Surpasses GMMs in Acoustic Modeling," ISCSLP2012, unpublished.
5. D. Yu, L. Deng, and G. Dahl, "Roles of Pre-training and Fine-Tuning in Context-Dependent DNN-HMMs for Real-World Speech Recognition," Proc. of NIPS Workshop on Deep Learning and Unsupervised Feature Learning, December 2010.
6. F. Seide, G. Li, X. Chen and D. Yu, "Feature Engineering in Context-Dependent Deep Neural Networks for Conversational Speech Transcription," Proc. ASRU 2011, pp. 24-29, 2011.
7. D. Yu, F. Seide, G. Li, L. Deng, "Exploiting sparseness in deep neural networks for large vocabulary speech recognition," in Proc. ICASSP 2012, pp. 4409-4412, 2012.
8. G. E. Hinton and R. R. Salakhutdinov, "Reducing the dimensionality of data with neural networks," Science, vol. 313, no. 5786, pp. 504-507, 2006.
9. Hongbing Hu, Stephen A. Zahorian, "Dimensionality reduction methods for HMM phonetic recognition," in Proc. ICASSP 2010, pp. 4854-4857, 2010.
10. D. Yu and M. L. Seltzer, "Improved Bottleneck Features Using Pretrained Deep Neural Networks," in Proc. Interspeech 2011, pp. 237-240, 2011.
11. V. Peter, P. Matus, J. Jozef, "Dimension reduction with principal component analysis applied to speech supervectors," Journal of Electrical and Electronics Engineering, vol. 4, no. 1, pp. 245-250, 2011.
12. P. C. Woodland, T. Hain, G. Evermann and D. Povey, "CU-HTK March 2001 Hub5 system." Proc. of DARPA Hub5E Conversational Speech Recognition Workshop, 2001.