Similar to large-language models in NLP, diffusion models have transformed generative modelling in computer vision and other multi-modal domains. The increasing interest in diffusion models is largely due to their simplicity and effectiveness. Anyone can train these models within a few lines of code, and their image quality and diversity are extremely good.
In this post, I list all papers on diffusion models that I could find on arxiv (using arxivscraper). I hope this list will be helpful in tracking exploding number of papers in this area. I intend to keep it updated (and further reduce false positives). If you find any paper missing, please let me know.
Cumulative number of papers on diffusion models over years. We include papers on both denoising-diffusion and score-based generative models. No doubt that the interest in this area is exploding in the academic community.