Training-Free Constrained Generation With Stable Diffusion Models
This addresses the need for strictly enforcing physical and functional constraints in generative models for applications like material design and content generation, representing a novel method for a known bottleneck.
The paper tackled the problem of incorporating strict domain-specific constraints into stable diffusion models for generative tasks, proposing a training-free integration with constrained optimization frameworks, and demonstrated effectiveness in material design and copyright-constrained generation with adherence to precise properties.
Stable diffusion models represent the state-of-the-art in data synthesis across diverse domains and hold transformative potential for applications in science and engineering, e.g., by facilitating the discovery of novel solutions and simulating systems that are computationally intractable to model explicitly. While there is increasing effort to incorporate physics-based constraints into generative models, existing techniques are either limited in their applicability to latent diffusion frameworks or lack the capability to strictly enforce domain-specific constraints. To address this limitation this paper proposes a novel integration of stable diffusion models with constrained optimization frameworks, enabling the generation of outputs satisfying stringent physical and functional requirements. The effectiveness of this approach is demonstrated through material design experiments requiring adherence to precise morphometric properties, challenging inverse design tasks involving the generation of materials inducing specific stress-strain responses, and copyright-constrained content generation tasks. All code has been released at https://github.com/RAISELab-atUVA/Constrained-Stable-Diffusion.