Post-Generation Curation of Synthetic Images via Homogeneous-Heterogeneous Splitting

Researchers introduced a post-generation curation method for synthetic images that selects informative subsets from a fixed pool of generated data. The approach splits each real class into homogeneous and heterogeneous subsets and scores synthetic images by a fidelity-diversity criterion, outperforming state-of-the-art data selection baselines and matching real-data performance with up to 40% fewer samples. The method is generator-agnostic and improves utility for classification and segmentation tasks.

arXiv:2607.02637v1 Announce Type: new Abstract: Recent generative models can produce high-quality synthetic images, offering scalable training training data for data-hungry models. Existing approaches to exploiting this potential typically involve 1 training or fine-tuning generators, or 2 using lightweight post-hoc adaptation like prompt engineering or inference-time guidance, making them generator-specific and expertise-intensive. We study a complementary question: given a fixed pool of generated images, can downstream utility be improved purely by selecting an informative subset? The answer is yes. We show that effective selection must counter a structural bias of modern generators: they tend to over-produce canonical modes of each class while underrepresenting intra-class variation. Building on this insight, we split each real class into a canonical Homogeneous HO subset and a non-redundant Heterogeneous HE subset, then score synthetic images by a fidelity-diversity criterion that rewards semantic alignment while penalizing canonical redundancy. The method is generator-agnostic and requires no retraining. Across multiple benchmarks, it consistently outperforms state-of-the-art data selection baselines and matches the real-data performance with up to 40% fewer synthetic samples. The same criterion remains effective when applied on top of stronger task-tuned generators, with gains on both classification and segmentation tasks. Post-generation selection is therefore not a substitute for better generators, but a complementary mechanism for improving the utility of synthetic data.