Investigating Nuisances in DCNN-based Face Recognition
Published in IEEE Transactions on Image Processing (TIP), 2018
In this paper, we present a thorough analysis of several aspects that impact on the use of DCNN for face recognition. The evaluation has been carried out from two main perspectives: the network architecture and the similarity measures used to compare deeply learned features; and the data (source and quality) and their preprocessing (bounding box and alignment). The results obtained on the IARPA Janus Benchmark-A, MegaFace, UMDFaces, and YouTube Faces data sets indicate viable hints for designing, training, and testing DCNNs. Download paper here
If you find our work useful, please cite us:
@article{ferrari2018investigating,
title={Investigating nuisances in DCNN-based face recognition},
author={Ferrari, Claudio and Lisanti, Giuseppe and Berretti, Stefano and Del Bimbo, Alberto},
journal={IEEE Transactions on Image Processing},
volume={27},
number={11},
pages={5638–5651},
year={2018},
publisher={IEEE}}