Projected Coupled Diffusion for Test-Time Constrained Joint Generation
This addresses a challenging test-time extension for diffusion models in applications like image-pair generation and multi-robot motion planning, though it appears incremental as it builds on existing test-time sampling methods.
The paper tackles the problem of generating jointly correlated samples from multiple pre-trained diffusion models while enforcing task-specific constraints without retraining, proposing Projected Coupled Diffusion (PCD) which shows improved coupling effects and guaranteed constraint satisfaction with low computational costs.
Modifications to test-time sampling have emerged as an important extension to diffusion algorithms, with the goal of biasing the generative process to achieve a given objective without having to retrain the entire diffusion model. However, generating jointly correlated samples from multiple pre-trained diffusion models while simultaneously enforcing task-specific constraints without costly retraining has remained challenging. To this end, we propose Projected Coupled Diffusion (PCD), a novel test-time framework for constrained joint generation. PCD introduces a coupled guidance term into the generative dynamics to encourage coordination between diffusion models and incorporates a projection step at each diffusion step to enforce hard constraints. Empirically, we demonstrate the effectiveness of PCD in application scenarios of image-pair generation, object manipulation, and multi-robot motion planning. Our results show improved coupling effects and guaranteed constraint satisfaction without incurring excessive computational costs.