Feature-wise transformations

Vincent Dumoulin (1) , Ethan Perez (2) , Harm Vries (1) , Florian Strub (3, 4) , Nathan Schucher (5) , Aaron Courville (1) , Yoshua Bengio (1)

1 UdeM - Université de Montréal
2 Rice University [Houston]
3 CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189
4 SEQUEL - Sequential Learning
5 Element AI

Many real-world problems require integrating multiple sources of information. Sometimes these problems involve multiple, distinct modalities of information — vision, language, audio, etc. — as is required to understand a scene in a movie or answer a question about an image. Other times, these problems involve multiple sources of the same kind of input, i.e. when summarizing several documents or drawing one image in the style of another.

Domaines: Réseau de neurones [cs.NE]; Intelligence artificielle [cs.AI]

Vincent Dumoulin, Ethan Perez, Harm Vries, Florian Strub, Nathan Schucher, et al.. Feature-wise transformations: A simple and surprisingly effective family of conditioning mechanisms.. Distill, 2018, 3 (7), ⟨10.23915/distill.00011⟩. ⟨hal-01841985⟩