Are Large-scale Datasets Necessary for Self-Supervised Pre-training?
Alaaeldin El-Nouby, Gautier Izacard, Hugo Touvron, Ivan Laptev, Hervé Jégou, Edouard Grave
People
Hervé Jégou
Research Lead
I have been a research scientist with Facebook AI Research since May 2015.
I am a former student of the Ecole Normale Superieure de Cachan, holding an MS (2002) and PhD (2005) from University of Rennes I. During my PhD, I worked on error-resilient compression and joint source channel coding. After that, I turned out to Computer Vision and Pattern Recognition. I joined the LEAR group (INRIA Grenoble) as a permanent researcher in 2006. I moved to INRIA Rennes in 2009, where I was involved or leading several projects related to large image and video collections. I am on leave from Inria since May 2015.
Interests
Computer vision
Latest Publications
arXiv - 2021
Gradient-based Adversarial Attacks against Text Transformers
Chuan Guo, Alexandre Sablayrolles, Hervé Jégou, Douwe Kiela
Training data-efficient image transformers & distillation through attention
Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, Hervé Jégou
Emerging Properties in Self-Supervised Vision Transformers
Mathilde Caron, Hugo Touvron, Ishan Misra, Hervé Jégou, Julien Mairal, Piotr Bojanowski, Armand Joulin
Going deeper with Image Transformers
Hugo Touvron, Matthieu Cord, Alexandre Sablayrolles, Gabriel Synnaeve, Hervé Jégou
Grafit: Learning fine-grained image representations with coarse labels
Hugo Touvron, Alexandre Sablayrolles, Matthijs Douze, Matthieu Cord, Hervé Jégou
