Single-Shot Domain Adaptation via Target-Aware Generative Augmentation
This addresses the challenge of adapting models to new domains with minimal data, which is crucial for real-world applications where data collection is limited, though it is incremental as it builds on existing adaptation techniques.
The paper tackles the problem of single-shot domain adaptation for deep neural networks by proposing SiSTA, a method that fine-tunes a generative model with one target sample and uses novel sampling to create synthetic data, achieving up to 20% improvement over baselines in face attribute detection and competitive performance with oracle models.
The problem of adapting models from a source domain using data from any target domain of interest has gained prominence, thanks to the brittle generalization in deep neural networks. While several test-time adaptation techniques have emerged, they typically rely on synthetic data augmentations in cases of limited target data availability. In this paper, we consider the challenging setting of single-shot adaptation and explore the design of augmentation strategies. We argue that augmentations utilized by existing methods are insufficient to handle large distribution shifts, and hence propose a new approach SiSTA (Single-Shot Target Augmentations), which first fine-tunes a generative model from the source domain using a single-shot target, and then employs novel sampling strategies for curating synthetic target data. Using experiments with a state-of-the-art domain adaptation method, we find that SiSTA produces improvements as high as 20\% over existing baselines under challenging shifts in face attribute detection, and that it performs competitively to oracle models obtained by training on a larger target dataset.