Difference between revisions of "Publications/bounthong.15.seminar"
From LRDE
(Created page with "{{CSIReport | authors = Jean-Luc Bounthong | title = Speaker specific i-vector channel compensation in speaker recognition | year = 2015 | abstract = The i-vector is actually ...") |
|||
| Line 3: | Line 3: | ||
| title = Speaker specific i-vector channel compensation in speaker recognition |
| title = Speaker specific i-vector channel compensation in speaker recognition |
||
| year = 2015 |
| year = 2015 |
||
| − | | abstract = The i-vector is actually the state of the art in speaker verification. Efficient result was achieved using classifier such as Cosine Distance (CD). Howeverclassification is performed on a global channel compensated i-vector. In this study, we explore the possibility to enroll a speaker and define a speaker specific channel compensation using i-vector. The objective is to improve the classifier performance using our previous work on Self-Organizing |
+ | | abstract = The i-vector is actually the state of the art in speaker verification. Efficient result was achieved using classifier such as Cosine Distance (CD). Howeverclassification is performed on a global channel compensated i-vector. In this study, we explore the possibility to enroll a speaker and define a speaker specific channel compensation using i-vector. The objective is to improve the classifier performance using our previous work on Self-Organizing Map to select suitable i-vector. We will compare the performance of our solution with the global channel compensated method. |
| type = techreport |
| type = techreport |
||
| id = bounthong.15.seminar |
| id = bounthong.15.seminar |
||
Latest revision as of 12:07, 15 May 2020
- Authors
- Jean-Luc Bounthong
- Type
- techreport
- Year
- 2015
Abstract
The i-vector is actually the state of the art in speaker verification. Efficient result was achieved using classifier such as Cosine Distance (CD). Howeverclassification is performed on a global channel compensated i-vector. In this study, we explore the possibility to enroll a speaker and define a speaker specific channel compensation using i-vector. The objective is to improve the classifier performance using our previous work on Self-Organizing Map to select suitable i-vector. We will compare the performance of our solution with the global channel compensated method.