AdAM: Few-Shot Image Generation via Adaptation-Aware Kernel Modulation
This work improves FSIG for scenarios where source and target domains are not closely related, which is incremental but addresses a specific bottleneck in existing methods.
The paper tackles the problem of few-shot image generation (FSIG) by addressing limitations in existing methods that fail to consider target domain adaptation in knowledge preservation, leading to poor performance when source and target domains are not closely related. It proposes Adaptation-Aware kernel Modulation (AdAM), which consistently achieves state-of-the-art performance across various FSIG setups, including challenging cases with more distant domains.
Few-shot image generation (FSIG) aims to learn to generate new and diverse images given few (e.g., 10) training samples. Recent work has addressed FSIG by leveraging a GAN pre-trained on a large-scale source domain and adapting it to the target domain with few target samples. Central to recent FSIG methods are knowledge preservation criteria, which select and preserve a subset of source knowledge to the adapted model. However, a major limitation of existing methods is that their knowledge preserving criteria consider only source domain/task and fail to consider target domain/adaptation in selecting source knowledge, casting doubt on their suitability for setups of different proximity between source and target domain. Our work makes two contributions. Firstly, we revisit recent FSIG works and their experiments. We reveal that under setups which assumption of close proximity between source and target domains is relaxed, many existing state-of-the-art (SOTA) methods which consider only source domain in knowledge preserving perform no better than a baseline method. As our second contribution, we propose Adaptation-Aware kernel Modulation (AdAM) for general FSIG of different source-target domain proximity. Extensive experiments show that AdAM consistently achieves SOTA performance in FSIG, including challenging setups where source and target domains are more apart.