Speech recognition: Acoustic, phonetic and lexical
Abstract
Our long-term research goal is the development and implementation of speaker-independent continuous speech recognition systems. It is our conviction that proper utilization of speech-specific knowledge is essential for advanced speech recognition systems. With this in mind, we have continued to make progress on the acquisition of acoustic-phonetic and lexical knowledge. We have completed the development of a continuous digit recognition system. The system was constructed to investigate the utilization of acoustic phonetic knowledge in a speech recognition system. Some of the significant development of this study includes a soft-failure procedure for lexical access, and the discovery of a set of acoustic-phonetic features for verification. We have completed a study of the constraints provided by lexical stress on word recognition. We found that lexical stress information alone can, on the average, reduce the number of word candidates from a large dictionary by more than 80%. In conjunction with this study, we successfully developed a system that automatically determines the stress pattern of a word from the acoustic signal.
- Publication:
-
NASA STI/Recon Technical Report N
- Pub Date:
- October 1985
- Bibcode:
- 1985STIN...8618589Z
- Keywords:
-
- Algorithms;
- Auditory Signals;
- Pattern Recognition;
- Phonetics;
- Sensory Discrimination;
- Speech Recognition;
- Acoustic Properties;
- Phonemics;
- Sound Waves;
- Vowels;
- Words (Language);
- Communications and Radar