Pascal Vincent
Research scientist in Meta's FAIR team; first author of the stacked denoising autoencoders paper
- Research Scientist, Fundamental AI Research (FAIR), Meta (FAIR)21
Pascal Vincent is a research scientist in the Fundamental AI Research (FAIR) team at Meta and a founding member of Mila – Quebec Artificial Intelligence Institute. Vincent's research on representation learning includes denoising autoencoders, described in Mila's biography as a precursor to self-supervised learning, and the denoising score matching principle, which the same biography notes is now used to train diffusion-based generative models. Vincent is first author of the 2010 JMLR paper on stacked denoising autoencoders and a co-author of I-JEPA, a self-supervised method that learns image representations by predicting the representations of target regions from a single context region.1234
Work
I-JEPA (Image Joint-Embedding Predictive Architecture) (external site: arxiv.org)Published research
Co-author of "Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture" (2023), which is also listed on Vincent's Meta AI profile.42
Stacked denoising autoencoders (external site: jmlr.org)Published research
First author of "Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion" (JMLR, 2010).31
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.