Contrastive Fine-grained Class Clustering via Generative Adversarial Networks
This addresses the problem of clustering subtle object details in images for computer vision applications, representing an incremental improvement over existing methods.
The paper tackled unsupervised fine-grained class clustering by introducing C3-GAN, which combines InfoGAN with contrastive learning to learn feature representations that form distinct cluster boundaries, achieving state-of-the-art performance on four fine-grained image datasets.
Unsupervised fine-grained class clustering is a practical yet challenging task due to the difficulty of feature representations learning of subtle object details. We introduce C3-GAN, a method that leverages the categorical inference power of InfoGAN with contrastive learning. We aim to learn feature representations that encourage a dataset to form distinct cluster boundaries in the embedding space, while also maximizing the mutual information between the latent code and its image observation. Our approach is to train a discriminator, which is also used for inferring clusters, to optimize the contrastive loss, where image-latent pairs that maximize the mutual information are considered as positive pairs and the rest as negative pairs. Specifically, we map the input of a generator, which was sampled from the categorical distribution, to the embedding space of the discriminator and let them act as a cluster centroid. In this way, C3-GAN succeeded in learning a clustering-friendly embedding space where each cluster is distinctively separable. Experimental results show that C3-GAN achieved the state-of-the-art clustering performance on four fine-grained image datasets, while also alleviating the mode collapse phenomenon. Code is available at https://github.com/naver-ai/c3-gan.