Loading timeline…
20202020s
Jonathan Ho
Lead author of the definitive paper establishing Denoising Diffusion Probabilistic Models (DDPM).
Organizations
UC BerkeleyGoogle BrainOpenAI
Major Achievements
- •His 2020 paper brought diffusion models from an obscure mathematical curiosity to state-of-the-art generation.
- •Proved they could generate significantly higher quality images than GANs.
- •The architecture he outlined became the absolute foundation for DALL-E, Midjourney, and Sora.
Key Papers
- Denoising Diffusion Probabilistic Models
Jonathan Ho, Ajay Jain, Pieter Abbeel · NeurIPS 2020
- Classifier-Free Diffusion Guidance
Jonathan Ho, Tim Salimans · NeurIPS 2021 Workshop · 2022