Few-shot Image Generation via Cross-domain Correspondence
This addresses the challenge of generating high-quality images with very limited data, which is incremental as it builds on existing GAN-based approaches.
The paper tackles the problem of overfitting in few-shot image generation by transferring diversity from a large source domain to a target domain with limited examples, achieving more diverse and realistic images than previous methods.
Training generative models, such as GANs, on a target domain containing limited examples (e.g., 10) can easily result in overfitting. In this work, we seek to utilize a large source domain for pretraining and transfer the diversity information from source to target. We propose to preserve the relative similarities and differences between instances in the source via a novel cross-domain distance consistency loss. To further reduce overfitting, we present an anchor-based strategy to encourage different levels of realism over different regions in the latent space. With extensive results in both photorealistic and non-photorealistic domains, we demonstrate qualitatively and quantitatively that our few-shot model automatically discovers correspondences between source and target domains and generates more diverse and realistic images than previous methods.