Joint Debiased Representation and Image Clustering Learning with Self-Supervision
This addresses the challenge of learning meaningful representations for minority classes in imbalanced datasets, which is an incremental improvement for computer vision applications.
The paper tackles the problem of joint clustering and contrastive learning on long-tailed data distributions, where existing methods fail due to majority classes overwhelming minority classes, and proposes a novel framework with a modified debiased contrastive loss and divergence clustering loss, showing improved performance across multiple datasets and tasks.
Contrastive learning is among the most successful methods for visual representation learning, and its performance can be further improved by jointly performing clustering on the learned representations. However, existing methods for joint clustering and contrastive learning do not perform well on long-tailed data distributions, as majority classes overwhelm and distort the loss of minority classes, thus preventing meaningful representations to be learned. Motivated by this, we develop a novel joint clustering and contrastive learning framework by adapting the debiased contrastive loss to avoid under-clustering minority classes of imbalanced datasets. We show that our proposed modified debiased contrastive loss and divergence clustering loss improves the performance across multiple datasets and learning tasks. The source code is available at https://anonymous.4open.science/r/SSL-debiased-clustering