It is well known that humans are far better in recalling person's face than his name. One of the promising approaches for devising a secure password scheme would be using a set of images as passwords rather than conventional PIN numbers. We have investigated such potential using an experimental paradigm by which security and usability for three different categories of images (i.e. landscapes scenes, random faces, familiar faces) can be compared. The results suggest that performance of the subjects was reliably higher for the "familiar faces" case than for other cases. Issues such as "Known-face attack" and "Camera Attack" were discussed. We propose a more secure, and yet usable, visual password system by exploiting human's innate capability of fast face identification against serial images of faces and implement it on a PDA for using within the pervasive and mobile environment.
|Number of pages
|Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
|Published - 2004
All Science Journal Classification (ASJC) codes
- Theoretical Computer Science
- General Computer Science