등록
CD-DNN-HMM 알고리즘 기반의 대어휘 연속 음성인식 시스템의 성능향상 방법
- 발명자
-
이성주, 정훈, 이윤근, 강병옥, 박전규
- 출원번호
-
15042309 (2016.02.12)
- 공개번호
-
20160240190 (2016.08.18)
- 등록번호
- 9805716 (2017.10.31)
- 출원국
- 미국
- 협약과제
-
14MS1500, 모바일 플랫폼 기반 대화모델 적용 자연어 음성인터페이스 기술 개발,
이윤근
- 초록
- Provided is an apparatus for large vocabulary continuous speech recognition (LVCSR) based on a context-dependent deep neural network hidden Markov model (CD-DNN-HMM) algorithm. The apparatus may include an extractor configured to extract acoustic model-state level information corresponding to an input speech signal from a training data model set using at least one of a first feature vector based on a gammatone filterbank signal analysis algorithm and a second feature vector based on a bottleneck algorithm, and a speech recognizer configured to provide a result of recognizing the input speech signal based on the extracted acoustic model-state level information.
- KSP 제안 키워드
- CD-DNN-HMM, Continuous Speech, Continuous Speech Recognition, DNN-HMM, Data Model, Deep neural network(DNN), Feature Vector, Gammatone filterbank, Hidden Markov, Large vocabulary, Markov model, Model set, Neural networks, Signal analysis, Speech Signals, Vector based, acoustic model, context-dependent, hidden Markov Model, large vocabulary continuous speech recognition, signal based, speech recognition, state level, training data
- 패밀리
-