Supervised Contrastive Learning
release_bqqi4sf4hfbfpgiiohomyjhsqu
by
Prannay Khosla, Piotr Teterwak, Chen Wang, Aaron Sarna, Yonglong Tian, Phillip Isola, Aaron Maschinot, Ce Liu, Dilip Krishnan
2020
Abstract
Contrastive learning applied to self-supervised representation learning has
seen a resurgence in recent years, leading to state of the art performance in
the unsupervised training of deep image models. Modern batch contrastive
approaches subsume or significantly outperform traditional contrastive losses
such as triplet, max-margin and the N-pairs loss. In this work, we extend the
self-supervised batch contrastive approach to the fully-supervised setting,
allowing us to effectively leverage label information. Clusters of points
belonging to the same class are pulled together in embedding space, while
simultaneously pushing apart clusters of samples from different classes. We
analyze two possible versions of the supervised contrastive (SupCon) loss,
identifying the best-performing formulation of the loss. On ResNet-200, we
achieve top-1 accuracy of 81.4% on the ImageNet dataset, which is 0.8% above
the best number reported for this architecture. We show consistent
outperformance over cross-entropy on other datasets and two ResNet variants.
The loss shows benefits for robustness to natural corruptions and is more
stable to hyperparameter settings such as optimizers and data augmentations.
Our loss function is simple to implement, and reference TensorFlow code is
released at https://t.ly/supcon.
In text/plain
format
Archived Files and Locations
application/pdf 2.2 MB
file_27cafsawa5ae3hfb3mcrgbn4my
|
arxiv.org (repository) web.archive.org (webarchive) |
2004.11362v3
access all versions, variants, and formats of this works (eg, pre-prints)