SMCL: Saliency Masked Contrastive Learning for Long-tailed Recognition
This work addresses the challenge of training models on imbalanced, long-tailed data, which is common in real-world applications, by mitigating bias towards majority classes.
The paper tackles the problem of long-tailed recognition by addressing how background features in imbalanced data lead to biased predictions towards majority classes, and proposes a method using saliency masking and contrastive learning to improve model generalizability, achieving state-of-the-art performance on benchmark datasets.
Real-world data often follow a long-tailed distribution with a high imbalance in the number of samples between classes. The problem with training from imbalanced data is that some background features, common to all classes, can be unobserved in classes with scarce samples. As a result, this background correlates to biased predictions into ``major" classes. In this paper, we propose saliency masked contrastive learning, a new method that uses saliency masking and contrastive learning to mitigate the problem and improve the generalizability of a model. Our key idea is to mask the important part of an image using saliency detection and use contrastive learning to move the masked image towards minor classes in the feature space, so that background features present in the masked image are no longer correlated with the original class. Experiment results show that our method achieves state-of-the-art level performance on benchmark long-tailed datasets.