Unconstrained Face Verification using Deep CNN Features

In this Iwo; we present an algorithm for unconstrained face verification based on deep convolutional features and evaluate it on the newly released IARPA Janus Benchmark A (LIB-A) dataset as well as on the traditional Labeled Face in the Wild (LFW) dataset. The IJB-A dataset includes real world unconstrained faces from 500 subjects with:hill pose and illumination variations which are much harder than the LFW and Youtube Face (YTF) datasets. The deep convolutional neural network (DCNN) is trained using the CASIAWebFace dataset. Results of experimental evaluations on the LIB-A and the LEW datasets are provided.