Title | ||
---|---|---|
Creating speaker independent ASR system through prosody modification based data augmentation |
Abstract | ||
---|---|---|
•Developing automatic speech recognition (ASR) system that are least affected by speaker-dependent variations.•Prosody-modification-based data augmentation is explored for creating Speaker Independent ASR system.•Prosody factors such as pitch and duration is explored to get pattern variations for training Speaker Independent ASR.•Relative improvements of 11:5% and 27% are obtained on decoding both adults’ and children’s test sets, respectively. |
Year | DOI | Venue |
---|---|---|
2020 | 10.1016/j.patrec.2019.12.019 | Pattern Recognition Letters |
DocType | Volume | ISSN |
Journal | 131 | 0167-8655 |
Citations | PageRank | References |
3 | 0.50 | 0 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
S. Shahnawazuddin | 1 | 64 | 17.34 |
Nagaraj Adiga | 2 | 12 | 4.45 |
Hemant Kumar Kathania | 3 | 19 | 4.27 |
B Tarun Sai | 4 | 3 | 0.50 |