Results 11 to 20 of about 18,616,519 (314)
Equivariant Diffusion for Molecule Generation in 3D [PDF]
International Conference on Machine Learning, 2022 This work introduces a diffusion model for molecule generation in 3D that is equivariant to Euclidean transformations. Our E(3) Equivariant Diffusion Model (EDM) learns to denoise a diffusion process with an equivariant network that jointly operates on ...
Emiel Hoogeboom +3 more
semanticscholar +1 more source
eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers [PDF]
arXiv.org, 2022 Large-scale diffusion-based generative models have led to breakthroughs in text-conditioned high-resolution image synthesis. Starting from random noise, such text-to-image diffusion models gradually synthesize images in an iterative fashion while ...
Y. Balaji +12 more
semanticscholar +1 more source
Diffusion Models for Adversarial Purification [PDF]
International Conference on Machine Learning, 2022 Adversarial purification refers to a class of defense methods that remove adversarial perturbations using a generative model. These methods do not make assumptions on the form of attack and the classification model, and thus can defend pre-existing ...
Weili Nie +5 more
semanticscholar +1 more source
GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation [PDF]
International Conference on Learning Representations, 2022 Predicting molecular conformations from molecular graphs is a fundamental problem in cheminformatics and drug discovery. Recently, significant progress has been achieved with machine learning approaches, especially with deep generative models.
Minkai Xu +5 more
semanticscholar +1 more source
Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models [PDF]
Computer Vision and Pattern Recognition, 2022 Text-conditioned image generation models have recently achieved astonishing results in image quality and text alignment and are consequently employed in a fast-growing number of applications.
P. Schramowski +3 more
semanticscholar +1 more source
DiGress: Discrete Denoising diffusion for graph generation [PDF]
International Conference on Learning Representations, 2022 This work introduces DiGress, a discrete denoising diffusion model for generating graphs with categorical node and edge attributes. Our model utilizes a discrete diffusion process that progressively edits graphs with noise, through the process of adding ...
Clément Vignac +5 more
semanticscholar +1 more source
Generalizing the Hypergraph Laplacian via a Diffusion Process with Mediators [PDF]
International Computing and Combinatorics Conference, 2018 In a recent breakthrough STOC 2015 paper, a continuous diffusion process was considered on hypergraphs (which has been refined in a recent JACM 2018 paper) to define a Laplacian operator, whose spectral properties satisfy the celebrated Cheeger’s ...
T-H. Hubert Chan, Zhibin Liang
semanticscholar +1 more source
Fast Sampling of Diffusion Models with Exponential Integrator [PDF]
International Conference on Learning Representations, 2022 The past few years have witnessed the great success of Diffusion models~(DMs) in generating high-fidelity samples in generative modeling tasks. A major limitation of the DM is its notoriously slow sampling procedure which normally requires hundreds to ...
Qinsheng Zhang, Yongxin Chen
semanticscholar +1 more source
Blended Diffusion for Text-driven Editing of Natural Images [PDF]
Computer Vision and Pattern Recognition, 2021 Natural language offers a highly intuitive interface for image editing. In this paper, we introduce the first solution for performing local (region-based) edits in generic natural images, based on a natural language description along with an ROI mask. We
Omri Avrahami, D. Lischinski, Ohad Fried
semanticscholar +1 more source
Training-Free Structured Diffusion Guidance for Compositional Text-to-Image Synthesis [PDF]
International Conference on Learning Representations, 2022 Large-scale diffusion models have achieved state-of-the-art results on text-to-image synthesis (T2I) tasks. Despite their ability to generate high-quality yet creative images, we observe that attribution-binding and compositional capabilities are still ...
Weixi Feng +8 more
semanticscholar +1 more source