讨论的论文题目: "Exploring Flow and Diffusion-based Generative Models with Scalable Interpolant Transformers." 场景: 一个会议厅。Dr. Evelyn Heun 站在讲台上,Professor Leonhard Euler-Maruyama 坐在她旁边。Dr. Ada ViT 在她的左边,Mr. James DiT 在她的右边。舞台上有一块大屏幕,显示论文的摘要和关键方...
We present Scalable Interpolant Transformers (SiT), a family of generative models built on the backbone of Diffusion Transformers (DiT). The interpolant framework, which allows for connecting two distributions in a more flexible way than standard diffusion models, makes possible a modular study of ...
一、Diffusion Probabilistic Models (DPMs) Diffusion-based generative models: forward/diffusion process:图中从右往左. 从x0经过好多个不同的q(xt|xt−1)到xT,相当于 VAE 中的 encoder. reverse process:图中从左往右. 从xT经过好多个不同的q(xt−1|xt)到x0,相当于 VAE 中的 decoder. (用pθ(xt...
Deep generative diffusion models are a promising avenue for 3D de novo molecular design in materials science and drug discovery. However, their utility is still limited by suboptimal performance on large molecular structures and limited training data. To address this gap, we explore the design space...
Abstract: We argue that the theory and practice of diffusion-based generative models are currently unnecessarily convoluted and seek to remedy the situation by presenting a design space that clearly separates the concrete design choices. This lets us identify several changes to both the sampling and ...
In this work, we attempt to tackle this constraint through a generative perspective and model the relationship between audio and text as their joint probability p(candidates,query). To this end, we present a diffusion-based ATR framework (DiffATR), which models ATR as an iterative procedure ...
Recently, several methods have been proposed to apply diffusion-based generative models to protein–ligand complexes [34,35,36]. For instance, DiffDock [35] modeled the conformation of a ligand relative to a given protein with a diffusion-based generative model. The authors reported significant pe...
由此产生的方法被称为基于分数的生成建模或扩散建模,在图像合成、文本到语音生成、时间序列预测和点云生成等应用中取得了创纪录的性能,挑战了生成对抗网络(GANS)在其中许多任务上的长期主导地位。 此外,基于分数的生成模型特别适合于贝叶斯推理任务,如求解病态逆问题,在医学图像重建中的几个任务上表现出优越的性能。
we propose a generative approach based on the stochastic diffusion method, namely TSDiff, for prediction of TS geometries just from 2D molecular graphs. TSDiff outperforms the existing ML models with 3D geometries in terms of both accuracy and efficiency. Moreover, it enables to sample various ...
"Speech Enhancement and Dereverberation with Diffusion-Based Generative Models", IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 31, pp. 2351-2364, 2023. [3] Julius Richter, Yi-Chiao Wu, Steven Krenn, Simon Welker, Bunlong Lay, Shinji Watanabe, Alexander Richard, Timo ...