Unconstrained Face Verification using Deep CNN Features

About

In this paper, we present an algorithm for unconstrained face verification based on deep convolutional features and evaluate it on the newly released IARPA Janus Benchmark A (IJB-A) dataset. The IJB-A dataset includes real-world unconstrained faces from 500 subjects with full pose and illumination variations which are much harder than the traditional Labeled Face in the Wild (LFW) and Youtube Face (YTF) datasets. The deep convolutional neural network (DCNN) is trained using the CASIA-WebFace dataset. Extensive experiments on the IJB-A dataset are provided.

Jun-Cheng Chen, Vishal M. Patel, Rama Chellappa · 2015

Related benchmarks

Task	Dataset	Result
Face Verification	LFW	Mean Accuracy—	417
Face Verification	LFW (Labeled Faces in the Wild) unrestricted-labeled-outside-data protocol 14	Accuracy97.45	47
Face Search	IJB-A	Rank@190.3	44
Face Verification	IJB-A	TAR @ FAR=1%0.838	38
Face Verification	IJB-A (test)	TAR @ FAR=0.010.967	37
Face Identification	IJB-A (test)	Rank-190.3	30
Face Recognition	CS2	Rank-1 Accuracy89.1	21
Face Verification	IJB-A (10 folds average)	TAR @ FAR=0.0183.8	18
Face Recognition	IJB-A (test)	TAR @ FAR=0.0183.8	16
Closed-set Face Search	IJB-A (10 folds average)	Rank-10.903	9