A picture's worth a thousand (private) words: Hierarchical generation of coherent synthetic photo albums
Google Research
Researchers at Google developed a method to generate synthetic photo albums that maintain privacy through differential privacy while preserving thematic coherence across images. The approach uses a hierarchical text-based intermediary process where photos are first converted to captions and album summaries, these representations are privately fine-tuned using large language models, and then converted back to images via text-to-image generation. The method achieved high semantic similarity (measured by MAUVE scores) between real and synthetic albums when tested on the YFCC100M dataset of nearly 100 million Creative Commons images.
Why it matters
Generative AI
