Center Contrastive Loss for Metric Learning
This addresses the problem of inefficient contrastive pair sampling in metric learning for researchers and practitioners, though it is incremental as it builds on existing contrastive and classification methods.
The paper tackles the challenge of sampling effective contrastive pairs in metric learning by proposing Center Contrastive Loss, which uses a class-wise center bank to compare with query data points, achieving state-of-the-art performance and faster convergence with a ResNet50 network.
Contrastive learning is a major studied topic in metric learning. However, sampling effective contrastive pairs remains a challenge due to factors such as limited batch size, imbalanced data distribution, and the risk of overfitting. In this paper, we propose a novel metric learning function called Center Contrastive Loss, which maintains a class-wise center bank and compares the category centers with the query data points using a contrastive loss. The center bank is updated in real-time to boost model convergence without the need for well-designed sample mining. The category centers are well-optimized classification proxies to re-balance the supervisory signal of each class. Furthermore, the proposed loss combines the advantages of both contrastive and classification methods by reducing intra-class variations and enhancing inter-class differences to improve the discriminative power of embeddings. Our experimental results, as shown in Figure 1, demonstrate that a standard network (ResNet50) trained with our loss achieves state-of-the-art performance and faster convergence.