Winners of the 11th European Biometrics Research Award

On 20 September 2017 the European Association for Biometrics (EAB) awarded young researchers for their outstanding works in the area of biometrics. An international jury chose 3 candidates out of a broad range of submitted high quality papers to present their significant contribution in front of the jury, the EAB members and the public audience.

Part of the jury of the EAB Research Award 2017: (f.r.t.l.) Tom Kevenaar, Jean-Christophe Fondeur, Josef Kittler, Javier Ortega Garcia, Martin Olsen

• The European Biometrics Research Award 2017 has been given to Nanang Susyanto from University of Twente, the Netherlands for his thesis titled A Non-standard Method on Biometric Score Level Fusion via Parametric Copula. This research presents new method for combining dependent scores from two or more biometric recognition systems, which is based on copula models. This copula is used to handle dependence between scores. For a given objective performance measure in a realistic scenario, a resampling method for choosing the best copula pair is proposed. The candidate showed that the method outperforms the existing likelihood-ratio-based score level fusions on some public biometric databases from fingerprint, face, speaker, and video-based gait recognitions under some common objective performance measures: maximizing acceptance rate at fixed false acceptance rate and minimizing discrimination loss.
  
  Winner of the EAB Research Award 2017: Nanang Susyanto
• This years European Biometrics Industry Award goes to Zhenhua Feng from University of Surrey, United Kingdom for the thesis titled Dynamic Attention-controlled Cascaded Shape Regression for Robust Facial Landmark Detection in the Wild. He presented a new approach for robust facial landmark detection on unconstrained faces. The key innovation of the approach is the fault-tolerant mechanism, using fuzzy set sample weighting, for attention-controlled domain-specific model training. Experimental results obtained on challenging datasets demonstrate the merits over the state-of- the-art methods.