ProCreate, Don't Reproduce! Propulsive Energy Diffusion for Creative Generation
This addresses the issue of overfitting and lack of creativity in generative models for AI researchers and practitioners, representing an incremental improvement over existing methods.
The paper tackles the problem of low sample diversity and training data reproduction in diffusion-based image generative models by proposing ProCreate, a method that propels generated image embeddings away from reference embeddings, achieving the highest sample diversity and fidelity on the FSCG-8 dataset and effectively preventing replication in large-scale evaluations.
In this paper, we propose ProCreate, a simple and easy-to-implement method to improve sample diversity and creativity of diffusion-based image generative models and to prevent training data reproduction. ProCreate operates on a set of reference images and actively propels the generated image embedding away from the reference embeddings during the generation process. We propose FSCG-8 (Few-Shot Creative Generation 8), a few-shot creative generation dataset on eight different categories -- encompassing different concepts, styles, and settings -- in which ProCreate achieves the highest sample diversity and fidelity. Furthermore, we show that ProCreate is effective at preventing replicating training data in a large-scale evaluation using training text prompts. Code and FSCG-8 are available at https://github.com/Agentic-Learning-AI-Lab/procreate-diffusion-public. The project page is available at https://procreate-diffusion.github.io.