Multiclass non-Adversarial Image Synthesis, with Application to Classification from Very Small Sample
This work addresses the problem of generating high-quality, diverse synthetic images, especially when training data is scarce, which is a significant challenge for researchers and practitioners working with limited datasets.
This paper introduces Clustered Optimization of LAtent space (COLA), a non-adversarial image synthesis method. It generates diverse multi-class images without supervision, outperforming previous non-adversarial methods in image quality and diversity. When applied to data augmentation for small datasets, COLA surpasses state-of-the-art performance in small-sample classification on CIFAR-10, CIFAR-100, STL-10, and Tiny-ImageNet.
The generation of synthetic images is currently being dominated by Generative Adversarial Networks (GANs). Despite their outstanding success in generating realistic looking images, they still suffer from major drawbacks, including an unstable and highly sensitive training procedure, mode-collapse and mode-mixture, and dependency on large training sets. In this work we present a novel non-adversarial generative method - Clustered Optimization of LAtent space (COLA), which overcomes some of the limitations of GANs, and outperforms GANs when training data is scarce. In the full data regime, our method is capable of generating diverse multi-class images with no supervision, surpassing previous non-adversarial methods in terms of image quality and diversity. In the small-data regime, where only a small sample of labeled images is available for training with no access to additional unlabeled data, our results surpass state-of-the-art GAN models trained on the same amount of data. Finally, when utilizing our model to augment small datasets, we surpass the state-of-the-art performance in small-sample classification tasks on challenging datasets, including CIFAR-10, CIFAR-100, STL-10 and Tiny-ImageNet. A theoretical analysis supporting the essence of the method is presented.