Speech Recognition

Project information

Database TIMIT (630 speakers, 438 males and 192 females ) 10 sentences for each speaker, 3 s each. Subset of 100 speakers Training: First 8 sentences, Testing: Last 2 sentences Speaker-Specific Feature MFCC Modelling GMM-UBM MAP adaptation

1 Commit
1 Branch
0 Tags

Created on

November 09, 2018

Loading