Time to augment self-supervised visual representation learning

no code implementations • 27 Jul 2022 • Arthur Aubret, Markus Ernst, Céline Teulière, Jochen Triesch

Specifically, our analyses reveal that: 1) 3-D object manipulations drastically improve the learning of object categories; 2) viewing objects against changing backgrounds is important for learning to discard background-related information from the latent representation.

Contrastive Learning Object +2