Hervé Jégou

Hervé is a Research Scientist at Facebook AI Research, where he currently serves as Site Lead of the Paris laboratory. After attending the École normale supérieure de Cachan, he received a Ph.D. from the University of Rennes for his thesis on error-resilient compression and joint source channel coding. After that, he turned his attention to computer vision and pattern recognition. Hervé joined INRIA as a permanent researcher, where he was involved in leading several projects related to large image and video collections.

Hervé's Publications

May 03, 2020

RESEARCH

COMPUTER VISION

Equi-normalization of Neural Networks

Modern neural networks are over-parametrized. In particular, each rectified linear hidden unit can be modified by a multiplicative factor by adjusting input and output weights, without changing the rest of the network. Inspired by the…

Pierre Stock, Benjamin Graham, Rémi Gribonval, Hervé Jégou,

May 03, 2020

May 03, 2020

RESEARCH

COMPUTER VISION

LAMV: Learning to align and match videos with kernelized temporal layers

This paper considers a learnable approach for comparing and aligning videos. Our architecture builds upon and revisits temporal match kernels within neural networks: we propose a new temporal layer that finds temporal alignments by maximizing…

Lorenzo Baraldi, Matthijs Douze, Rita Cucchiara, Hervé Jégou,

May 03, 2020

May 03, 2020

RESEARCH

COMPUTER VISION

Link and code: Fast indexing with graphs and compact regression codes

Similarity search approaches based on graph walks have recently attained outstanding speed-accuracy trade-offs, taking aside the memory requirements. In this paper, we revisit these approaches by considering, additionally, the memory constraint…

Matthijs Douze, Alexandre Sablayrolles, Hervé Jégou,

May 03, 2020