Prosody dependent speech recognition on radio news corpus of American English

International Journal
2006-01-01 14:00
Authors : Ken Chen, Mark Hasegawa-johnson, Aaron Cohen, Sarah Borys, Sung-Suk Kim, Jennifer Cole, Jeung-Yoon Choi

Year : 2006

Publisher / Conference : IEEE Transactions on Audio, Speech, and Language Processing

Volume : 14, issue 1

Page : 232-245

Does prosody help word recognition? This paper proposes a novel probabilistic framework in which word and phoneme are dependent on prosody in a way that reduces word error rates (WER) relative to a prosody-independent recognizer with comparable parameter count. In the proposed prosody-dependent speech recognizer, word and phoneme models are conditioned on two important prosodic variables: the intonational phrase boundary and the pitch accent. An information-theoretic analysis is provided to show that prosody dependent acoustic and language modeling can increase the mutual information between the true word hypothesis and the acoustic observation by exciting the interaction between prosody dependent acoustic model and prosody dependent language model. Empirically, results indicate that the influence of these prosodic variables on allophonic models are mainly restricted to a small subset of distributions: the duration PDFs (modeled using an explicit duration hidden Markov model or EDHMM) and the acoustic-prosodic observation PDFs (normalized pitch frequency). Influence of prosody on cepstral features is limited to a subset of phonemes: for example, vowels may be influenced by both accent and phrase position, but phrase-initial and phrase-final consonants are independent of accent. Leveraging these results, effective prosody dependent allophonic models are built with minimal increase in parameter count. These prosody dependent speech recognizers are able to reduce word error rates by up to 11% relative to prosody independent recognizers with comparable parameter count, in experiments based on the prosodically-transcribed Boston Radio News corpus.
전체 356
46 Domestic Conference 이준호, 정재웅, 박영철, 윤대희 "홈 엔터네인먼트 시스템을 위한 효율적인 3차원 공간 등화 방법" in 2006년도 하계종합학술발표회, 2006
45 Domestic Conference 이규하, 황운철, 이충용, 윤대희 "다중채널 SDR 시스템을 위한 비대칭 폴리페이즈 필터 뱅크 기반의 채널화/역채널화 구조" in 2006년도 하계종합학술발표회, 2006
44 Domestic Journal 윤성완, 강홍구, 윤대희 "음성신호의 특성을 고려한 패킷 손실 은닉 알고리즘" in 한국통신학회논문지, vol.31, No.7C, pp.691-699, 2006
43 Domestic Journal 이창준,이재성,박영철 "MPEG 오디오 부호화기를 위한 향상된 프리 에코 컨트롤 알고리듬" in 방송공학회논문지, vol.11, 제 2호, pp.191-199, 2006
42 International Conference Yang-Won Jung, Hyen-O Oh, Hyo Jin Kim, Seung Jong Choi "New CLD Quantization Method for Spatial Audio Coding" in 120th Convention of Audio Engineering Society, pp.6734, 2006
41 International Conference Junho Lee, Jae-woong Jeong, Young-Cheol Park, she-Woong Jeong, Dae Hee Youn "Effective Room Equalization Based on Warped Common Acoustical Poles and Zeros" in 120th Convention of Audio Engineering Society, pp.6721, 2006
40 International Conference Kyoung Ho Bang, Young Cheol Park, Dae Hee Youn "A Dual Audio Transcoding Algorithm for Digital Multimedia Broadcasting Services" in 120th Convention of Audio Engineering Society, pp.6814, 2006
39 International Conference Tacksung Choi, Young-Cheol Park, Dae Hee Youn "Efficient Out of Head Localization System for Mobile Applications" in 120th Convention of Audio Engineering Society, pp.6758, 2006
38 International Conference Kyoung Ho Bang, Young Cheol Park, Dae Hee Youn "Audio Transcoding Algorithm For Mobile Multimedia Application" in ICASSP, 2006
37 International Conference Bong-Jin Lee, Sung-Wan Yoon, Hong-Goo Kang, Dae Hee Youn "On the Use of Voting Methods for Speaker Identification Based on Various Resolution Filterbanks" in ICASSP, 2006