Compression for 3D face recognition applications

This paper proposes a novel 3D lossy compression algorithm tailored for 3D faces. We analyse the effects of compression on the face verification rate and measure recognition performances on the Face Recognition Grand Challenge database. Whilst preserving the spatial resolution enabling reconstruction of surface details, the proposed scheme achieves substantial compression to the extent that personal 3D biometric data could fit on a 2D barcode.

• Type: Conference
• Year: 2007
• Where: AVSS
• Authors: Lorenzo Granai, Miroslav Hamouz, Jose Rafael Tena, Theodore Vlachos
• Keywords: AVSS 2007, Face Verification Rate, Lossy Compression Algorithm, Measure Recognition Performances, Signal Processing