A Speech Recognition IC Using Hidden Markov Models with Continuous Observation Densities
| Autor: | Kong-Pang Pun, Chiu-Sing Choy, Wei Han, Cheong-Fat Chan, Kwok-Wai Hon |
|---|---|
| Rok vydání: | 2007 |
| Předmět: |
Vocabulary
Computer science business.industry Speech recognition Gaussian media_common.quotation_subject Pattern recognition Viterbi algorithm Chip symbols.namesake Signal Processing Pattern recognition (psychology) symbols State (computer science) Artificial intelligence Electrical and Electronic Engineering Hidden Markov model business Information Systems media_common |
| Zdroj: | The Journal of VLSI Signal Processing Systems for Signal, Image, and Video Technology. 47:223-232 |
| ISSN: | 1573-109X 0922-5773 |
| DOI: | 10.1007/s11265-007-0049-6 |
| Popis: | This paper presents the design of a speech recognition IC using hidden Markov models (HMMs) with continuous observation densities. Results of offline and live recognition tests are also given. Our design employs a table look-up method to simplify the computation and hence the architecture of the circuit. Currently each state of the HMMs is represented by a double-mixture Gaussian distribution. With minor modifications, the proposed architecture can be extended to implement a recognizer in which models with higher order multi-mixture Gaussian distribution are used for more precise acoustic modeling. The test chip is fabricated with a 0.35 μm CMOS technology. The maximum operating frequency is 62.5 MHz at 3.3 V. For a 50-word vocabulary, the estimated recognition time is about 0.16 s. Using noise-corrupted utterances, the recognition accuracy is 93.8% for isolated English digits. Such a performance is comparable to the software implementation with the same algorithm. Live recognition test was also run for a vocabulary of 11 Chinese words. The accuracy is 91.8% for five male and five female speakers. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|