Bi-modal biometric authentication on mobile phones in challenging conditions

作者：

Highlights：

• We examine bi-modal (face/speaker) authentication in challenging mobile environment.

• We release new protocols (and data) with significant mismatch conditions (MOBIO).

• We study bi-modal and multi-algorithm fusion using generative modelling techniques.

• Multi-algorithm and multi-modal fusion provides a consistent performance improvement.

• The proposed bi-modal system significantly outperforms the state-of-the-art.

摘要

•We examine bi-modal (face/speaker) authentication in challenging mobile environment.•We release new protocols (and data) with significant mismatch conditions (MOBIO).•We study bi-modal and multi-algorithm fusion using generative modelling techniques.•Multi-algorithm and multi-modal fusion provides a consistent performance improvement.•The proposed bi-modal system significantly outperforms the state-of-the-art.

论文关键词：Face authentication,Speaker authentication,Bi-modal authentication,Gaussian mixture model,Session variability,Inter-session variability,Total variability,I-vector,Fusion

论文评审过程：Received 30 April 2013, Accepted 3 October 2013, Available online 14 October 2013.

论文官网地址：https://doi.org/10.1016/j.imavis.2013.10.001