12 papers
Ambient Diffusion Policy: Imitation Learning from Suboptimal Data in Robotics
Adam Wei, Nicholas Pfaff, Thomas Cohn +4
We propose Ambient Diffusion Policy, a simple and principled method for imitation learning from suboptimal data in robotics. High-quality, task-specific robot data is expensive and…
Align & Invert: Solving Inverse Problems with Diffusion and Flow-based Models via Representation Alignment
Loukas Sfountouris, Giannis Daras, Paris Giampouras
Enforcing alignment between the internal representations of diffusion or flow-based generative models and those of pretrained self-supervised encoders has recently been shown to pr…
Does Generation Require Memorization? Creative Diffusion Models using Ambient Diffusion
Kulin Shah, Alkis Kalavasis, Adam R. Klivans +1
There is strong empirical evidence that the state-of-the-art diffusion modeling paradigm leads to models that memorize the training set, especially when the training set is small.…
Ambient Physics: Training Neural PDE Solvers with Partial Observations
Harris Abdul Majid, Giannis Daras, Francesco Tudisco +1
In many scientific settings, acquiring complete observations of PDE coefficients and solutions can be expensive, hazardous, or impossible. Recent diffusion-based methods can recons…
Ambient Dataloops: Generative Models for Dataset Refinement
Adrián RodrÃguez-Muñoz, William Daspit, Adam Klivans +3
We propose Ambient Dataloops, an iterative framework for refining datasets that makes it easier for diffusion models to learn the underlying data distribution. Modern datasets cont…
DiffEM: Learning from Corrupted Data with Diffusion Models via Expectation Maximization
Danial Hosseintabar, Fan Chen, Giannis Daras +2
Diffusion models have emerged as powerful generative priors for high-dimensional inverse problems, yet learning them when only corrupted or noisy observations are available remains…