THE ALGORITHM FOR DETERMINING WORD BOUNDARIES IN CONTINUOUS SPEECH IN TERMS OF NOISE POLLUTION BASED ON THE ANALYSIS OF THE ENTROPY OF THE SPEECH SIGNAL
Automatic speech recognition, especially in noisy environments, is a complex task. The most important stage of automatic speech recognition is the correct definition of word boundaries in the speech stream. Even a slight improvement at the stage of highlighting the boundaries of words, significantly...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | Russian |
| Published: |
North-Caucasus Federal University
2022-10-01
|
| Series: | Современная наука и инновации |
| Subjects: | |
| Online Access: | https://msi.elpub.ru/jour/article/view/641 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Automatic speech recognition, especially in noisy environments, is a complex task. The most important stage of automatic speech recognition is the correct definition of word boundaries in the speech stream. Even a slight improvement at the stage of highlighting the boundaries of words, significantly affects the operation of the entire speech recognition system. To recognize isolated words, this problem boils down to determining the correct word boundary. For continuous speech, this task is much more difficult, since the speech signal is a continuous stream without any speech pauses. The most promising approach involves the use of speech signal entropy to search for word boundaries. |
|---|---|
| ISSN: | 2307-910X |