Subventions et des contributions :

Subvention ou bourse octroyée s'appliquant à plus d'un exercice financier (2017-2018 à 2018-2019).

A practical face verification system is expected to achieve a False Accept Rate of 0.01% or lower whilex000D
ensuring a False Reject Rate (FRR) of 5% or lower. In recent years, face recognition algorithms based on deepx000D
neural networks have achieved high accuracy when tested on face recognition database. However, when putx000D
into real-world application, those algorithms are often not robust enough, due to factors such as differentx000D
lighting conditions, camera distances, face orientations, and occlusions.x000D
As a startup company, AltumView Systems Inc. has been developing various intelligent camera products sincex000D
January 2015. In this project, the university team will work with AltumView to improve the performance andx000D
robustness of its existing face recognition products. Research to be conducted includes: 1) try differentx000D
configurations of the cost functions of the deep neural network scheme, and identify the most effective way; 2)x000D
design new neural network architecture that is robust to different lighting conditions, camera distances, andx000D
face angles etc. between the ID photos and wild input photos; 3) design schemes to deal with partially occludedx000D
faces caused by e.g., glasses; 4) investigate Generative Adversarial Networks for preprocessing of input photosx000D
and domain adaptation; and 5) investigate how to combine different techniques to improve the overallx000D
performance of the system.