English
Related papers

Related papers: Stitched Value Model for Diffusion Alignment

200 papers

Text-conditioned image generation models have recently shown immense qualitative success using denoising diffusion processes. However, unlike discriminative vision-and-language models, it is a non-trivial task to subject these…

Computer Vision and Pattern Recognition · Computer Science 2023-11-06 Benno Krojer , Elinor Poole-Dayan , Vikram Voleti , Christopher Pal , Siva Reddy

A fairly reliable trend in deep reinforcement learning is that the performance scales with the number of parameters, provided a complimentary scaling in amount of training data. As the appetite for large models increases, it is imperative…

Machine Learning · Computer Science 2023-06-14 Bogdan Mazoure , Walter Talbott , Miguel Angel Bautista , Devon Hjelm , Alexander Toshev , Josh Susskind

Reinforcement learning (RL) has improved guided image generation with diffusion models by directly optimizing rewards that capture image quality, aesthetics, and instruction following capabilities. However, the resulting generative policies…

Computer Vision and Pattern Recognition · Computer Science 2024-06-25 Owen Oertell , Jonathan D. Chang , Yiyi Zhang , Kianté Brantley , Wen Sun

Despite the remarkable progress in deep generative models, synthesizing high-resolution and temporally coherent videos still remains a challenge due to their high-dimensionality and complex temporal dynamics along with large spatial…

Computer Vision and Pattern Recognition · Computer Science 2023-03-31 Sihyun Yu , Kihyuk Sohn , Subin Kim , Jinwoo Shin

Large pretrained plain vision Transformers (ViTs) have been the workhorse for many downstream tasks. However, existing works utilizing off-the-shelf ViTs are inefficient in terms of training and deployment, because adopting ViTs with…

Computer Vision and Pattern Recognition · Computer Science 2023-11-29 Zizheng Pan , Jing Liu , Haoyu He , Jianfei Cai , Bohan Zhuang

Image-conditioned Video diffusion models achieve impressive visual realism but often suffer from weakened motion fidelity, e.g., reduced motion dynamics or degraded long-term temporal coherence, especially after fine-tuning. We study the…

Computer Vision and Pattern Recognition · Computer Science 2026-03-19 Xi Ye , Wenjia Yang , Yangyang Xu , Xiaoyang Liu , Duo Su , Mengfei Xia , Jun Zhu

Diffusion models have recently shown promise in offline RL. However, these methods often suffer from high training costs and slow convergence, particularly when using transformer-based denoising backbones. While several optimization…

Machine Learning · Computer Science 2025-06-23 Zhiying Qiu , Tao Lin

Distilling latent diffusion models (LDMs) into ones that are fast to sample from is attracting growing research interest. However, the majority of existing methods face two critical challenges: (1) They hinge on long training using a huge…

Computer Vision and Pattern Recognition · Computer Science 2024-11-08 Qingsong Xie , Zhenyi Liao , Zhijie Deng , Chen chen , Haonan Lu

Video diffusion alignment has been heavily relied on scalar rewards. These rewards are typically derived from learned reward models in human preference datasets, requiring additional training and extensive collection. Moreover, scalar…

Computer Vision and Pattern Recognition · Computer Science 2026-03-18 Yifan Wang , Yanyu Li , Gordon Guocheng Qian , Sergey Tulyakov , Yun Fu , Anil Kag

Diffusion models have achieved impressive success in generating photorealistic images, but challenges remain in ensuring precise semantic alignment with input prompts. Optimizing the initial noisy latent offers a more efficient alternative…

Computer Vision and Pattern Recognition · Computer Science 2025-10-28 Boming Miao , Chunxiao Li , Xiaoxiao Wang , Andi Zhang , Rui Sun , Zizhe Wang , Yao Zhu

Reward-based fine-tuning steers a pretrained diffusion or flow-based generative model toward higher-reward samples while remaining close to the pretrained model. Although existing methods are derived from different perspectives, we show…

Machine Learning · Computer Science 2026-05-08 Jeongjae Lee , Jinho Chang , Jeongsol Kim , Jong Chul Ye

Denoising diffusion models have been a mainstream approach for image generation, however, training these models often suffers from slow convergence. In this paper, we discovered that the slow convergence is partly due to conflicting…

Computer Vision and Pattern Recognition · Computer Science 2024-03-12 Tiankai Hang , Shuyang Gu , Chen Li , Jianmin Bao , Dong Chen , Han Hu , Xin Geng , Baining Guo

Aligning diffusion models with user preferences has been a key challenge. Existing methods for aligning diffusion models either require retraining or are limited to differentiable reward functions. To address these limitations, we propose a…

Computer Vision and Pattern Recognition · Computer Science 2025-02-28 Po-Hung Yeh , Kuang-Huei Lee , Jun-Cheng Chen

Diffusion and flow matching models have achieved remarkable success in text-to-image generation. However, these models typically rely on the predetermined denoising schedules for all prompts. The multi-step reverse diffusion process can be…

Computer Vision and Pattern Recognition · Computer Science 2025-03-06 Zilyu Ye , Zhiyang Chen , Tiancheng Li , Zemin Huang , Weijian Luo , Guo-Jun Qi

Diffusion models have demonstrated a remarkable ability in Text-to-Image (T2I) generation applications. Despite the advanced generation output, they suffer from heavy computation overhead, especially for large models that contain tens of…

Computer Vision and Pattern Recognition · Computer Science 2026-03-10 Desen Sun , Jason Hon , Jintao Zhang , Sihang Liu

Flow and diffusion models produce high-quality samples, but adapting them to user preferences or constraints post-training remains costly and brittle, a challenge commonly called reward alignment. We argue that efficient reward alignment…

Diffusion models achieve state-of-the-art generative performance but are fundamentally bottlenecked by their slow, iterative sampling process. While diffusion distillation techniques enable high-fidelity, few-step generation, traditional…

Computer Vision and Pattern Recognition · Computer Science 2026-04-01 Linqian Fan , Peiqin Sun , Tiancheng Wen , Shun Lu , Chengru Song

This tutorial provides an in-depth guide on inference-time guidance and alignment methods for optimizing downstream reward functions in diffusion models. While diffusion models are renowned for their generative modeling capabilities,…

Artificial Intelligence · Computer Science 2025-01-22 Masatoshi Uehara , Yulai Zhao , Chenyu Wang , Xiner Li , Aviv Regev , Sergey Levine , Tommaso Biancalani

The modeling, computational cost, and accuracy of traditional Spatio-temporal networks are the three most concentrated research topics in video action recognition. The traditional 2D convolution has a low computational cost, but it cannot…

Computer Vision and Pattern Recognition · Computer Science 2021-12-07 Zhaoqilin Yang , Gaoyun An

Diffusion models, such as Stable Diffusion, have shown incredible performance on text-to-image generation. Since text-to-image generation often requires models to generate visual concepts with fine-grained details and attributes specified…

Computer Vision and Pattern Recognition · Computer Science 2024-04-26 Xuehai He , Weixi Feng , Tsu-Jui Fu , Varun Jampani , Arjun Akula , Pradyumna Narayana , Sugato Basu , William Yang Wang , Xin Eric Wang