Abstract
This paper proposes to use combined acoustic and visual feature vectors to distinguish live synchronous audio-video recordings from replay attacks that use audio with a still photo. Equal error rates below 2 % are achieved using a multi-dimensional eigenlip representation and EERs of 7% are achieved with a one-dimensional lip-opening ratio.
| Original language | English |
|---|---|
| Title of host publication | Proceedings Interspeech 2004 |
| Editors | Soon Hyob Kim, Dae Hee Youn |
| Place of Publication | Jeju, Korea |
| Publisher | Interspeech 2004 |
| Pages | 2509-2512 |
| Number of pages | 4 |
| ISBN (Print) | 1225-441X |
| Publication status | Published - 2004 |
| Event | INTERSPEECH 2004 - ICSLP 8th International Conference on Spoken Language Processing - Jeju, Korea, Republic of Duration: 3 Oct 2004 → 7 Oct 2004 |
Conference
| Conference | INTERSPEECH 2004 - ICSLP 8th International Conference on Spoken Language Processing |
|---|---|
| Country/Territory | Korea, Republic of |
| City | Jeju |
| Period | 3/10/04 → 7/10/04 |