The quality of input samples is a crucial issue for both verification and identification biometric systems. Literature offers many interesting hints in some contexts, like fingerprint matching, but it is still deficient in others, like face matching. This paper proposes new quality indices for face samples which are based either on the measure of pose or illumination distortion, or on the measure of face symmetry. Results show that they give better results than other proposals in literature.
Measuring measures for face sample quality / Maria De, Marsico; Michele, Nappi; Riccio, Daniel. - (2011), pp. 7-12. (Intervento presentato al convegno 3rd international ACM workshop on Multimedia in forensics and intelligence tenutosi a Scottsdale (AZ), USA nel 28/11-01/12/2011) [10.1145/2072521.2072524].
Measuring measures for face sample quality
RICCIO, Daniel
2011
Abstract
The quality of input samples is a crucial issue for both verification and identification biometric systems. Literature offers many interesting hints in some contexts, like fingerprint matching, but it is still deficient in others, like face matching. This paper proposes new quality indices for face samples which are based either on the measure of pose or illumination distortion, or on the measure of face symmetry. Results show that they give better results than other proposals in literature.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
