English
Related papers

Related papers: U-REPA: Aligning Diffusion U-Nets to ViTs

200 papers

Representation Alignment (REPA) has emerged as a simple way to accelerate Diffusion Transformers training in latent space. At the same time, pixel-space diffusion transformers such as Just image Transformers (JiT) have attracted growing…

Computer Vision and Pattern Recognition · Computer Science 2026-03-17 Jaeyo Shin , Jiwook Kim , Hyunjung Shim

Recent advances in Diffusion Transformers (DiTs) demonstrate that aligning noisy latent states with well-trained semantic features-as pioneered by Representation Alignment (REPA)-can substantially accelerate training and improve generation…

Computer Vision and Pattern Recognition · Computer Science 2026-05-19 Shaodong Xu , Zhendong Wang , Litong Gong , Zexian Li , Wengang Zhou , Tiezheng Ge , Houqiang Li

Recent studies have shown that the denoising process in (generative) diffusion models can induce meaningful (discriminative) representations inside the model, though the quality of these representations still lags behind those learned…

Computer Vision and Pattern Recognition · Computer Science 2025-06-19 Sihyun Yu , Sangkyung Kwak , Huiwon Jang , Jongheon Jeong , Jonathan Huang , Jinwoo Shin , Saining Xie

Diffusion Transformers (DiTs) deliver state-of-the-art image quality, yet their training remains notoriously slow. A recent remedy -- representation alignment (REPA) that matches DiT hidden features to those of a non-generative teacher…

Computer Vision and Pattern Recognition · Computer Science 2025-05-23 Ziqiao Wang , Wangbo Zhao , Yuhao Zhou , Zekai Li , Zhiyuan Liang , Mingjia Shi , Xuanlei Zhao , Pengfei Zhou , Kaipeng Zhang , Zhangyang Wang , Kai Wang , Yang You

Fine-tuning Video Diffusion Models (VDMs) at the user level to generate videos that reflect specific attributes of training data presents notable challenges, yet remains underexplored despite its practical importance. Meanwhile, recent work…

Computer Vision and Pattern Recognition · Computer Science 2025-06-26 Sungwon Hwang , Hyojin Jang , Kinam Kim , Minho Park , Jaegul Choo

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 provide a powerful inductive bias, improving both convergence…

Computer Vision and Pattern Recognition · Computer Science 2026-05-27 Loukas Sfountouris , Giannis Daras , Paris Giampouras

Physics-informed diffusion models typically enforce PDE constraints only on final outputs, leaving intermediate representations unconstrained and prone to shortcut learning under shifted boundary conditions. We introduce **REPA-P**, a…

Machine Learning · Computer Science 2026-05-21 Haozhe Jia , Pengyu Yin , Wenshuo Chen , Shaofeng Liang , Lei Wang , Bowen Tian , Xiucheng Wang , Nanqian Jia , Yutao Yue

Representation alignment (REPA) guides generative training by distilling representations from a strong, pretrained vision encoder to intermediate diffusion features. We investigate a fundamental question: what aspect of the target…

Computer Vision and Pattern Recognition · Computer Science 2025-12-12 Jaskirat Singh , Xingjian Leng , Zongze Wu , Liang Zheng , Richard Zhang , Eli Shechtman , Saining Xie

Representation alignment has recently emerged as an effective paradigm for accelerating Diffusion Transformer training. Despite their success, existing alignment methods typically impose a fixed supervision target or a fixed alignment…

Computer Vision and Pattern Recognition · Computer Science 2026-05-06 Ruibin Min , Yexin Liu , Aimin Pan , Changsheng Lu , Jiafei Wu , Kelu Yao , Xiaogang Xu , Harry Yang

Diffusion Transformers (DiTs) introduce the transformer architecture to diffusion tasks for latent-space image generation. With an isotropic architecture that chains a series of transformer blocks, DiTs demonstrate competitive performance…

Computer Vision and Pattern Recognition · Computer Science 2024-10-31 Yuchuan Tian , Zhijun Tu , Hanting Chen , Jie Hu , Chao Xu , Yunhe Wang

Diffusion-based image compression has recently shown outstanding perceptual fidelity, yet its practicality is hindered by prohibitive sampling overhead and high memory usage. Most existing diffusion codecs employ U-Net architectures, where…

Image and Video Processing · Electrical Eng. & Systems 2026-03-16 Junqi Shi , Ming Lu , Xingchen Li , Anle Ke , Ruiqi Zhang , Zhan Ma

While recent advancements in generative modeling have significantly improved text-image alignment, some residual misalignment between text and image representations still remains. Some approaches address this issue by fine-tuning models in…

Computer Vision and Pattern Recognition · Computer Science 2025-12-11 Jaa-Yeon Lee , Byunghee Cha , Jeongsol Kim , Jong Chul Ye

Diffusion-based methods have been acknowledged as a powerful paradigm for end-to-end visuomotor control in robotics. Most existing approaches adopt a Diffusion Policy in U-Net architecture (DP-U), which, while effective, suffers from…

Robotics · Computer Science 2025-09-30 Linzhi Wu , Aoran Mei , Xiyue Wang , Guo-Niu Zhu , Zhongxue Gan

REPA and its variants effectively mitigate training challenges in diffusion models by incorporating external visual representations from pretrained models, through alignment between the noisy hidden projections of denoising networks and…

Computer Vision and Pattern Recognition · Computer Science 2025-09-30 Ge Wu , Shen Zhang , Ruijing Shi , Shanghua Gao , Zhenyuan Chen , Lei Wang , Zhaowei Chen , Hongcheng Gao , Yao Tang , Jian Yang , Ming-Ming Cheng , Xiang Li

Unified Multimodal models (UMMs) built on a single architecture have shown impressive performance in both understanding and generation. We identify a fundamental challenge that lies in inductive biases induced by distinct supervision…

Computer Vision and Pattern Recognition · Computer Science 2026-05-26 Renjie Lu , Xulong Zhang , Xiaoyang Qu , Shangfei Wang , Jianzong Wang

REPresentation Alignment (REPA) improves the training of generative flow models by aligning intermediate hidden states with pretrained teacher features, but its effectiveness in token-conditioned audio Flow Matching critically depends on…

Sound · Computer Science 2026-05-29 Pengfei Zhang , Tianxin Xie , Minghao Yang , Li Liu

Contrastive pre-training on image-text pairs, exemplified by CLIP, becomes a standard technique for learning multi-modal visual-language representations. Although CLIP has demonstrated remarkable performance, training it from scratch on…

Computer Vision and Pattern Recognition · Computer Science 2024-06-17 Jihao Liu , Jinliang Zheng , Boxiao Liu , Yu Liu , Hongsheng Li

Recent studies have highlighted the interplay between diffusion models and representation learning. Intermediate representations from diffusion models can be leveraged for downstream visual tasks, while self-supervised vision models can…

Computer Vision and Pattern Recognition · Computer Science 2025-07-23 Xiangxiang Chu , Renda Li , Yong Wang

Diffusion models can be improved with additional guidance towards more effective representations of input. Indeed, prior empirical work has already shown that aligning internal representations of the diffusion model with those of…

Machine Learning · Computer Science 2025-10-14 Chenyu Wang , Cai Zhou , Sharut Gupta , Zongyu Lin , Stefanie Jegelka , Stephen Bates , Tommi Jaakkola

While representation alignment with self-supervised models has been shown to improve diffusion model training, its potential for enhancing inference-time conditioning remains largely unexplored. We introduce Representation-Aligned Guidance…

Computer Vision and Pattern Recognition · Computer Science 2026-02-04 Nicolas Sereyjol-Garros , Ellington Kirby , Victor Letzelter , Victor Besnier , Nermin Samet
‹ Prev 1 2 3 10 Next ›