@inproceedings{900c10074f54486c8b71d35970ca3edd,
title = "ProCreate, Don{\textquoteright}t Reproduce! Propulsive Energy Diffusion for Creative Generation",
abstract = "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.",
keywords = "AI-assisted design, data replication, few-shot generation, Generative models",
author = "Jack Lu and Ryan Teehan and Mengye Ren",
note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.; 18th European Conference on Computer Vision, ECCV 2024 ; Conference date: 29-09-2024 Through 04-10-2024",
year = "2025",
doi = "10.1007/978-3-031-73027-6_23",
language = "English (US)",
isbn = "9783031730269",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Science and Business Media Deutschland GmbH",
pages = "397--414",
editor = "Ale{\v s} Leonardis and Elisa Ricci and Stefan Roth and Olga Russakovsky and Torsten Sattler and G{\"u}l Varol",
booktitle = "Computer Vision – ECCV 2024 - 18th European Conference, Proceedings",
address = "Germany",
}