YangLing0818/Diffusion-Models-Papers-Survey-Taxonomy
A comprehensive collection and categorization of diffusion model papers, accompanied by a detailed survey and taxonomy.
Core Features
Detailed Introduction
This repository is a meticulously curated knowledge hub for diffusion models, built upon a comprehensive survey accepted by ACM Computing Surveys. It systematically collects and categorizes research papers, offering a detailed taxonomy of algorithms, including sampling acceleration, likelihood maximization, and handling diverse data structures, as well as integration with LLMs and DPO/RLHF. The project also outlines a broad spectrum of applications from computer vision to molecular modeling and explores connections with other generative models. Designed to be a continuously updated resource, it provides researchers and practitioners with an organized and current overview of the rapidly evolving diffusion model landscape.