Generalized Parametric Contrastive Learning
This addresses the challenge of imbalanced data in machine learning, particularly for computer vision tasks like recognition and segmentation, offering a novel optimization-based solution that is incremental but effective.
The paper tackles the problem of class imbalance in contrastive learning by proposing Generalized Parametric Contrastive Learning (GPaCo/PaCo), which introduces learnable class centers to rebalance optimization and achieves state-of-the-art results on long-tailed benchmarks, with models showing better generalization and robustness on ImageNet and improvements in semantic segmentation tasks.
In this paper, we propose the Generalized Parametric Contrastive Learning (GPaCo/PaCo) which works well on both imbalanced and balanced data. Based on theoretical analysis, we observe that supervised contrastive loss tends to bias high-frequency classes and thus increases the difficulty of imbalanced learning. We introduce a set of parametric class-wise learnable centers to rebalance from an optimization perspective. Further, we analyze our GPaCo/PaCo loss under a balanced setting. Our analysis demonstrates that GPaCo/PaCo can adaptively enhance the intensity of pushing samples of the same class close as more samples are pulled together with their corresponding centers and benefit hard example learning. Experiments on long-tailed benchmarks manifest the new state-of-the-art for long-tailed recognition. On full ImageNet, models from CNNs to vision transformers trained with GPaCo loss show better generalization performance and stronger robustness compared with MAE models. Moreover, GPaCo can be applied to the semantic segmentation task and obvious improvements are observed on the 4 most popular benchmarks. Our code is available at https://github.com/dvlab-research/Parametric-Contrastive-Learning.