Conditional Diffusion with Less Explicit Guidance via Model Predictive Control

Anonymous

Keywords: conditional diffusion, conditional sampling, model predictive control, guidance

TL;DR: Improved conditional diffusion sampling quality at just 5 time steps with conditional guidance

Abstract: How much explicit guidance is necessary for conditional diffusion? We consider the problem of conditional sampling using an unconditional diffusion model and limited explicit guidance (e.g., a noised classifier, or a conditional diffusion model) that is restricted to a small number of time steps. We explore a model predictive control (MPC)-like approach to approximate guidance by simulating unconditional diffusion forward, and backpropagating explicit guidance feedback. MPC-approximated guides have high cosine similarity to real guides, even over large simulation distances. Adding MPC steps improves generative quality when explicit guidance is limited to five time steps

Student Paper: No