Authors
Iacopo Masi*, Anh Tuan Tran*, Tal Hassner*, Jatuporn Toy Leksut, Gerard Medioni
Publication date
2016
Conference
European Conference on Computer Vision Workshops (ECCV)
Description
Face recognition capabilities have recently made extraordinary leaps. Though this progress is at least partially due to ballooning training set sizes – huge numbers of face images downloaded and labeled for identity – it is not clear if the formidable task of collecting so many images is truly necessary. We propose a far more accessible means of increasing training data sizes for face recognition systems: Domain specific data augmentation. We describe novel methods of enriching an existing dataset with important facial appearance variations by manipulating the faces it contains. This synthesis is also used when matching query images represented by standard convolutional neural networks. The effect of training and testing with synthesized images is tested on the LFW and IJB-A (verification and identification) benchmarks and Janus CS2. The performances obtained by our approach match state of the art …
Scholar articles
I Masi, AT Trần, T Hassner, JT Leksut, G Medioni - Computer Vision–ECCV 2016: 14th European …, 2016
I Masi, AT Tran, T Hassner, JT Leksut, G Medioni - 2016
I Masi, FJ Chang, J Choi, S Harel, J Kim, KG Kim - IEEE Trans. Pattern Anal. Mach. Intell, 2019
I Masi, AT Tran - Do we really need to collect millions of faces for …, 2016
M Iacopo, ATTT Hassner, LJ Toy, M Gerard - European Conf. on Computer Vision, 2016
MITAT Hassner, TLJTM Gérard - Computer Vision–ECCV, 2016
I Masi - Do we really need to collect millions of faces for …