Few-shot Image Generation via Cross-domain Correspondence

Computer Vision and Pattern Recognition (CVPR'21)

Published June 23, 2021

Utkarsh Ojha, Yijun Li, Jingwan (Cynthia) Lu, Alexei A. Efros, Yong Jae Lee, Eli Shechtman, Richard Zhang

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.

Learn More