Generating High Fidelity Data from Low-density Regions using Diffusion Models

Citations Over TimeTop 10% of 2022 papers

Abstract

Our work focuses on addressing sample deficiency from low-density regions of data manifold in common image datasets. We leverage diffusion process based generative models to synthesize novel images from low-density regions. We observe that uniform sampling from diffusion models predominantly samples from high-density regions of the data manifold. Therefore, we modify the sampling process to guide it towards low-density regions while simulta-neously maintaining the fidelity of synthetic data. We rigorously demonstrate that our process successfully generates novel high fidelity samples from low-density regions. We further examine generated samples and show that the model does not memorize low-density data and indeed learns to generate novel samples from low-density regions.

Related Papers

• → Initial sampling using multi-fidelity information in simulation optimization of manufacturing systems(2018)1 cited
• → Towards Understanding the Interplay of Generative Artificial Intelligence and the Internet(2023)9 cited
• → Generative Model for Person Re-Identification: A Review(2020)
• → Falsification of Learning-Based Controllers through Multi-Fidelity Bayesian Optimization(2022)
• → TC-VAE: Uncovering Out-of-Distribution Data Generative Factors(2023)