Timezone: »

 
Poster
Unsupervised Learning of Visual Features by Contrasting Cluster Assignments
Mathilde Caron · Ishan Misra · Julien Mairal · Priya Goyal · Piotr Bojanowski · Armand Joulin

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1698
in Poster Session 6 »

Unsupervised image representations have significantly reduced the gap with supervised pretraining, notably with the recent achievements of contrastive learning methods. These contrastive methods typically work online and rely on a large number of explicit pairwise feature comparisons, which is computationally challenging. In this paper, we propose an online algorithm, SwAV, that takes advantage of contrastive methods without requiring to compute pairwise comparisons. Specifically, our method simultaneously clusters the data while enforcing consistency between cluster assignments produced for different augmentations (or views) of the same image, instead of comparing features directly as in contrastive learning. Simply put, we use a swapped prediction mechanism where we predict the code of a view from the representation of another view. Our method can be trained with large and small batches and can scale to unlimited amounts of data. Compared to previous contrastive methods, our method is more memory efficient since it does not require a large memory bank or a special momentum network. In addition, we also propose a new data augmentation strategy, multi-crop, that uses a mix of views with different resolutions in place of two full-resolution views, without increasing the memory or compute requirements. We validate our findings by achieving 75.3% top-1 accuracy on ImageNet with ResNet-50, as well as surpassing supervised pretraining on all the considered transfer tasks.

Author Information

Mathilde Caron (INRIA / FAIR)
Ishan Misra (Facebook AI Research )
Julien Mairal (Inria)
Priya Goyal (Facebook AI Research)
Piotr Bojanowski (Facebook)
Armand Joulin (Facebook AI research)

More from the Same Authors