Grounding inductive biases in natural images: invariance stems from variations in data
Diane Bouchacourt, Mark Ibrahim, Ari S. Morcos
People
Diane Bouchacourt
Research Scientist
I am a research scientist at Facebook Artificial Intelligence Research (FAIR), starting in November 2017. Prior to this, I pursued a PhD in Machine Learning in the Optimization for Vision and Learning (OVAL) Group at the University of Oxford, UK. My PhD focused on learning probability distributions of structured objects, with applications in computer vision, and was funded by the Microsoft Research PhD Scholarship Programme in EMEA.
Interests
Deep probabilistic models, optimization, structured prediction and unsupervised/weakly supervised learning
Latest Publications
NeurIPS - 2021
Entropy Minimization In Emergent Languages
Eugene Kharitonov, Rahma Chaabouni, Diane Bouchacourt, Marco Baroni
Compositionality and Generalization in Emergent Languages
Rahma Chaabouni, Eugene Kharitonov, Diane Bouchacourt, Emmanuel Dupoux, Marco Baroni
Permutation Equivariant Models for Compositional Generalization in Language
Jonathan Gordon, David Lopez-Paz, Marco Baroni, Diane Bouchacourt
EGG: a toolkit for research on Emergence of lanGuage in Games
Eugene Kharitonov, Rahma Chaabouni, Diane Bouchacourt, Marco Baroni
Miss Tools and Mr Fruit: Emergent communication in agents learning about object affordances
Diane Bouchacourt, Marco Baroni