Few-shot Image Generation with Mixup-based Distance Learning
This addresses the challenge of few-shot image generation for AI applications where data is scarce, though it is incremental as it enhances existing models rather than introducing a new paradigm.
The paper tackles the problem of training generative models like GANs with limited data, which often leads to overfitting and mode collapse, by proposing a mixup-based distance regularization method that improves fidelity and diversity in few-shot image synthesis.
Producing diverse and realistic images with generative models such as GANs typically requires large scale training with vast amount of images. GANs trained with limited data can easily memorize few training samples and display undesirable properties like "stairlike" latent space where interpolation in the latent space yields discontinuous transitions in the output space. In this work, we consider a challenging task of pretraining-free few-shot image synthesis, and seek to train existing generative models with minimal overfitting and mode collapse. We propose mixup-based distance regularization on the feature space of both a generator and the counterpart discriminator that encourages the two players to reason not only about the scarce observed data points but the relative distances in the feature space they reside. Qualitative and quantitative evaluation on diverse datasets demonstrates that our method is generally applicable to existing models to enhance both fidelity and diversity under few-shot setting. Code is available.