Spotlight Poster
TFG: Unified Training-Free Guidance for Diffusion Models
Haotian Ye · Haowei Lin · Jiaqi Han · Minkai Xu · Sheng Liu · Yitao Liang · Jianzhu Ma · James Zou · Stefano Ermon
East Exhibit Hall A-C #2506
Given an unconditional diffusion model and a predictor for a target property of interest (e.g., a classifier), the goal of training-free guidance is to generate samples with desirable target properties without additional training. Existing methods, though effective in various individual applications, often lack theoretical grounding and rigorous testing on extensive benchmarks. As a result, they could even fail on simple tasks, and applying them to a new problem becomes unavoidably difficult. This paper introduces a novel algorithmic framework encompassing existing methods as special cases, unifying the study of training-free guidance into the analysis of an algorithm-agnostic design space. Via theoretical and empirical investigation, we propose an efficient and effective hyper-parameter searching strategy that can be readily applied to any downstream task. We systematically benchmark across 7 diffusion models on 16 tasks with 40 targets, and improve performance by 8.5% on average. Our framework and benchmark offer a solid foundation for conditional generation in a training-free manner.
Live content is unavailable. Log in and register to view live content