Abstract | ||
---|---|---|
We developed a speaker verification system that is efficient for short utterances. The i-vector-based speaker representation has helped realize highly accurate speaker verification systems, however, it might be not robust against short utterances because the reliability of statistics required for extracting i-vectors is low. On the other hand, multiple kernel learning based on conditional entropy minimization has also achieved high accuracy in speaker verification that is robust against intra-speaker variability. To improve the robustness of speaker verification systems against short utterances, we attempted to integrate the above-mentioned complementary systems. Our experimental results showed that the proposed system integration achieved high-accuracy speaker verification systems, irrespective of the utterance lengths, even for very short utterances (e.g., less than two seconds). |
Year | DOI | Venue |
---|---|---|
2013 | 10.1109/ACPR.2013.42 | ACPR |
Keywords | Field | DocType |
high-accuracy speaker verification system,short utterances,proposed system integration,short utterance,accurate speaker verification system,i-vector-based speaker representation,speaker verification,speaker verification system,conditional entropy minimization,i-vector-based speaker verification,above-mentioned complementary system,learning artificial intelligence,speaker recognition,entropy | Computer science,Multiple kernel learning,Utterance,Robustness (computer science),Speech recognition,Minification,Speaker recognition,Speaker diarisation,Conditional entropy,System integration | Conference |
Citations | PageRank | References |
0 | 0.34 | 13 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hideitsu Hino | 1 | 99 | 25.73 |
Tetsuji Ogawa | 2 | 73 | 27.96 |