Equivariant Self-Supervised Learning Improves IT Predictivity

T Yerxa, J Feather, E P Simoncelli and SY Chung

Published in Annual Meeting, Cognitive Computational Neuroscience, Aug 2024.

Download:
  • Reprint (pdf)

  • We present a novel method for self-supervised learning of representations that are equivariant to a set of transfor- mations. When trained on images, we demonstrate that the learned representations effectively factorize sources of variability in their inputs, and provide improved pre- diction of responses of cells in macaque visual area IT across four different datasets.
  • Listing of all publications