Compositional Generative Inverse Design
This addresses inverse design challenges in fields like mechanical and aerospace engineering by providing a more robust and compositional method, though it appears incremental as it builds on existing diffusion model approaches.
The paper tackles the problem of inverse design by optimizing over learned energy functions from diffusion models to avoid adversarial modes, improving design performance and enabling compositional design of complex systems. In experiments, it generalizes to more objects in an N-body task and discovers formation flying to minimize drag in a multi-airfoil design task.
Inverse design, where we seek to design input variables in order to optimize an underlying objective function, is an important problem that arises across fields such as mechanical engineering to aerospace engineering. Inverse design is typically formulated as an optimization problem, with recent works leveraging optimization across learned dynamics models. However, as models are optimized they tend to fall into adversarial modes, preventing effective sampling. We illustrate that by instead optimizing over the learned energy function captured by the diffusion model, we can avoid such adversarial examples and significantly improve design performance. We further illustrate how such a design system is compositional, enabling us to combine multiple different diffusion models representing subcomponents of our desired system to design systems with every specified component. In an N-body interaction task and a challenging 2D multi-airfoil design task, we demonstrate that by composing the learned diffusion model at test time, our method allows us to design initial states and boundary shapes that are more complex than those in the training data. Our method generalizes to more objects for N-body dataset and discovers formation flying to minimize drag in the multi-airfoil design task. Project website and code can be found at https://github.com/AI4Science-WestlakeU/cindm.