扩散模型在机器人学研究中的最新进展与应用
本项目整理了扩散模型在机器人学领域的前沿研究文献和代码资源,涵盖模仿学习、视频生成等多个研究方向。资源库提供扩散模型入门教程、精选论文摘要和主题分类的机器人扩散论文集,为研究者提供全面参考。项目旨在帮助研究人员了解并应用机器人扩散技术的最新进展。
"Creating noise from data is easy; creating data from noise is generative modeling."
Yang Song in "Score-Based Generative Modeling through Stochastic Differential Equations" Song et al., 2020
This repository offers a brief summary of essential papers and blogs on diffusion models, alongside a categorized collection of robotics diffusion papers and useful code repositories for starting your own diffusion robotics project.
2.1 Imitation Learning and Policy Learning
2.2 Video Diffusion in Robotics
2.3 Online RL
2.4 Offline RL
2.5 Inverse RL
2.6 World Models
<a name="Learning-about-Diffusion-models"></a> While there exist many tutorials for Diffusion models, below you can find an overview of some of the best introduction blog posts and video:
What are Diffusion Models?: an introduction video, which introduces the general idea of diffusion models and some high-level math about how the model works
Diffusion Models | Paper Explanation | Math Explained another great video tutorial explaining the math and notation of diffusion models in detail with visual aid
Generative Modeling by Estimating Gradients of the Data Distribution: blog post from the one of the most influential authors in this area, which introduces diffusion models from the score-based perspective
What are Diffusion Models: a in-depth blog post about the theory of diffusion models with a general summary on how diffusion model improved over time
Understanding Diffusion Models: an in-depth explanation paper, which explains the diffusion models from both perspectives with detailed derivations
If you don't like reading blog posts and prefer the original papers, below you can find a list with the most important diffusion theory papers:
Sohl-Dickstein, Jascha, et al. "Deep unsupervised learning using nonequilibrium thermodynamics." International Conference on Machine Learning. PMLR, 2015.
Ho, Jonathan, et al. "Denoising diffusion probabilistic models." Advances in Neural Information Processing Systems 33 (2020): 6840-6851.
Song, Yang, et al. "Score-Based Generative Modeling through Stochastic Differential Equations." International Conference on Learning Representations. 2020.
Ho, Jonathan, and Tim Salimans. "Classifier-Free Diffusion Guidance." NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications. 2021.
Karras, Tero, et al. "Elucidating the Design Space of Diffusion-Based Generative Models." Advances in Neural Information Processing Systems 35 (2022)
A general list with all published diffusion papers can be found here: Whats the score?
<a name="Diffusion-in-Robotics"></a> Since the modern diffusion models have been around for only 3 years, the literature about diffusion models in the context of robotics is still small, but growing rapidly. Below you can find most robotics diffusion papers, which have been published at conferences or uploaded to Arxiv so far:
<a name="Imitation-Learning-and-Policy-Learning"></a>
Zhou, Hongyi, et al. "Variational Distillation of Diffusion Policies into Mixture of Experts." arXiv preprint arXiv:2406.12538 (2024).
Jia, Xiaogang, et al. "MaIL: Improving Imitation Learning with Mamba." arXiv preprint arXiv:2406.08234 (2024).
Hao, Ce, et al. "Language-Guided Manipulation with Diffusion Policies and Constrained Inpainting." arXiv preprint arXiv:2406.09767 (2024).
Shridhar, Mohit, Yat Long Lo, and Stephen James. "Generative Image as Action Models." arXiv preprint arXiv:2407.07875 (2024).
Høeg, Sigmund H., and Lars Tingelstad. "TEDi Policy: Temporally Entangled Diffusion for Robotic Control." arXiv preprint arXiv:2406.04806 (2024).
Vosylius, Vitalis, et al. "Render and Diffuse: Aligning Image and Action Spaces for Diffusion-based Behaviour Cloning." Proceedings of Robotics: Science and Systems (RSS) 2024.
Prasad, Aaditya, et al. "Consistency Policy: Accelerated Visuomotor Policies via Consistency Distillation." Proceedings of Robotics: Science and Systems (RSS) 2024.
Bharadhwaj, Homanga, et al. "Track2Act: Predicting Point Tracks from Internet Videos enables Diverse Zero-shot Robot Manipulation." ECCV 2024
Reuss, Moritz, et al. "Multimodal Diffusion Transformer: Learning Versatile Behavior from Multimodal Goals." Proceedings of Robotics: Science and Systems (RSS) 2024.
Gupta, Gunshi, et al. "Pre-trained Text-to-Image Diffusion Models Are Versatile Representation Learners for Control." First Workshop on Vision-Language Models for Navigation and Manipulation at ICRA 2024 (2024).
Ke, Tsung-Wei, Nikolaos Gkanatsios, and Katerina Fragkiadaki. "3D Diffuser Actor: Policy Diffusion with 3D Scene Representations." arXiv preprint arXiv:2402.10885 (2024).
Ze, Yanjie, et al. "3D Diffusion Policy." Proceedings of Robotics: Science and Systems (RSS) 2024.
Ma, Xiao, et al. "Hierarchical Diffusion Policy for Kinematics-Aware Multi-Task Robotic Manipulation." arXiv preprint arXiv:2403.03890 (2024).
Yan, Ge, Yueh-Hua Wu, and Xiaolong Wang. "DNAct: Diffusion Guided Multi-Task 3D Policy Learning." arXiv preprint arXiv:2403.04115 (2024).
Zhang, Xiaoyu, et al. "Diffusion Meets DAgger: Supercharging Eye-in-hand Imitation Learning." arXiv preprint arXiv:2402.17768 (2024).
Chen, Kaiqi, et al. "Behavioral Refinement via Interpolant-based Policy Diffusion." arXiv preprint arXiv:2402.16075 (2024).
Wang, Bingzheng, et al. "DiffAIL: Diffusion Adversarial Imitation Learning." arXiv preprint arXiv:2312.06348 (2023).
Scheikl, Paul Maria, et al. "Movement Primitive Diffusion: Learning Gentle Robotic Manipulation of Deformable Objects." arXiv preprint arXiv:2312.10008 (2023).
Octo Model Team et al. Octo: An Open-Source Generalist Robot Policy
Black, Kevin, et al. "ZERO-SHOT ROBOTIC MANIPULATION WITH PRETRAINED IMAGE-EDITING DIFFUSION MODELS." arXiv preprint arXiv:2310.10639 (2023).
Reuss, Moritz, and Rudolf Lioutikov. "Multimodal Diffusion Transformer for Learning from Play." 2nd Workshop on Language and Robot Learning: Language as Grounding. 2023.
Sridhar, Ajay, et al. "NoMaD: Goal Masked Diffusion Policies for Navigation and Exploration." arXiv preprint arXiv:2310.07896 (2023).
Zhou, Xian, et al. "Unifying Diffusion Models with Action Detection Transformers for Multi-task Robotic Manipulation." Conference on Robot Learning. PMLR, 2023.
Ze, Yanjie, et al. "Multi-task real robot learning with generalizable neural feature fields." 7th Annual Conference on Robot Learning. 2023.
Mishra, Utkarsh Aashu, et al. "Generative Skill Chaining: Long-Horizon Skill Planning with Diffusion Models." Conference on Robot Learning. PMLR, 2023.
Chen, Lili, et al. "PlayFusion: Skill Acquisition via Diffusion from Language-Annotated Play." Conference on Robot Learning. PMLR, 2023.
Ha, Huy, Pete Florence, and Shuran Song. "Scaling Up and Distilling Down: Language-Guided Robot Skill Acquisition." Conference on Robot Learning. PMLR, 2023.
Xu, Mengda, et al. "XSkill: Cross Embodiment Skill Discovery." Conference on Robot Learning. PMLR, 2023.
Li, Xiang, et al. "Crossway Diffusion: Improving Diffusion-based Visuomotor Policy via Self-supervised Learning." arXiv preprint arXiv:2307.01849 (2023).
Ng, Eley, Ziang Liu, and Monroe Kennedy III. "Diffusion Co-Policy for Synergistic Human-Robot Collaborative Tasks." arXiv preprint arXiv:2305.12171 (2023).
Chi, Cheng, et al. "Diffusion Policy: Visuomotor Policy Learning via Action Diffusion." Proceedings of Robotics: Science and Systems (RSS) 2023.
Reuss, Moritz, et al. "Goal-Conditioned Imitation Learning using Score-based Diffusion Policies." Proceedings of Robotics: Science and Systems (RSS) 2023.
Yoneda, Takuma, et al. "To the Noise and Back: Diffusion for Shared Autonomy." Proceedings of Robotics: Science and Systems (RSS) 2023.
Jiang, Chiyu, et al. "MotionDiffuser: Controllable Multi-Agent Motion Prediction Using Diffusion." Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2023.
Kapelyukh, Ivan, et al. "DALL-E-Bot: Introducing Web-Scale Diffusion Models to Robotics." IEEE Robotics and Automation Letters (RA-L) 2023.
Pearce, Tim, et al. "Imitating human behaviour with diffusion models." " International Conference on Learning Representations. 2023.
Yu, Tianhe, et al. "Scaling robot learning with semantically imagined experience." arXiv preprint arXiv:2302.11550 (2023).
<a name="Video-Diffusion"></a>
The ability of Diffusion models to generate realistic videos over a long horizon has enabled new applications in the context of robotics.
Wang, Boyang, et al. "This&That: Language-Gesture Controlled Video Generation for Robot Planning." arXiv:2407.05530 (2024).
Chen, Boyuan, et al. "Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion." arXiv preprint arXiv:2407.01392 (2024).
Zhou, Siyuan, et al. "RoboDreamer: Learning Compositional World Models for Robot Imagination." arXiv preprint arXiv:2404.12377 (2024).
McCarthy, Robert, et al. "Towards Generalist Robot Learning from Internet Video: A Survey." arXiv preprint arXiv:2404.19664 (2024).
He, Haoran, et al. "Large-Scale Actionless Video Pre-Training via Discrete Diffusion for Efficient Policy Learning." arXiv preprint arXiv:2402.14407 (2024).
Liang, Zhixuan, et al. "SkillDiffuser: Interpretable Hierarchical Planning via Skill Abstractions in Diffusion-Based Task Execution." arXiv preprint arXiv:2312.11598 (2023).
Huang, Tao, et al. "Diffusion Reward: Learning Rewards via Conditional Video Diffusion." arXiv preprint arXiv:2312.14134 (2023).
Du, Yilun, et al. "Video Language Planning." arXiv preprint arXiv:2310.10625 (2023).
Yang, Mengjiao, et al. "Learning Interactive Real-World Simulators." arXiv preprint arXiv:2310.06114 (2023).
Ko, Po-Chen, et al. "Learning to Act from Actionless Videos through Dense Correspondences." arXiv preprint arXiv:2310.08576 (2023).
Ajay, Anurag, et al. "Compositional Foundation Models for Hierarchical Planning." Advances in Neural Information Processing Systems 37 (2023)
Dai, Yilun, et al. "Learning Universal Policies via Text-Guided Video Generation." Advances in Neural Information Processing Systems 37 (2023)
AI Excel全自动制表工具
AEE 在线 AI 全自动 Excel 编辑器,提供智能录入、自动公式、数据整理、图表生成等功能,高效处理 Excel 任务,提升办公效率。支持自动高亮数据、批量计算、不规则数据录入,适用于企业、教育、金融等多场景。
基于 UI-TARS 视觉语言模型的桌面应用,可通过自然语言控制计算机进行多模态操作。
UI-TARS-desktop 是一款功能强大的桌面应用,基于 UI-TARS(视觉语言模型)构建。它具备自然语言控制、截图与视觉识别、精确的鼠标键盘控制等功能,支持跨平台使用(Windows/MacOS),能提供实时反馈和状态显示,且数据完全本地处理,保障隐私安全。该应用集成了多种大语言模型和搜索方式,还可进行文件系统操作。适用于需要智能交互和自动化任务的场景,如信息检索、文件管理等。其提供了详细的文档,包括快速启动、部署、贡献指南和 SDK 使用说明等,方便开发者使用和扩展。
开源且先进的大规模视频生成模型项目
Wan2.1 是一个开源且先进的大规模视频生成模型项目,支持文本到图像、文本到视频、图像到视频等多种生成任务。它具备丰富的配置选项,可调整分辨率、扩散步数等参数,还能对提示词进行增强。使用了多种先进技术和工具,在视频和图像生成领域具有广泛应用前景,适合研究人员和开发者使用。
全流程 AI 驱动的数据可视化工具,助力用户轻松创作高颜值图表
爱图表(aitubiao.com)就是AI图表,是由镝数科技推出的一款创新型智能数据可视化平台,专注于为用户提供便捷的图表生成、数据分析和报告撰写服务。爱图表是中国首个在图表场景接入DeepSeek的产品。通过接入前沿的DeepSeek系列AI模型,爱图表结合强大的数据处理能力与智能化功能,致力于帮助职场人士高效处理和表达数据,提升工作效率和报告质量。
一款强大的视觉语言模型,支持图像和视频输入
Qwen2.5-VL 是一款强大的视觉语言模型,支持图像和视频输入,可用于多种场景,如商品特点总结、图像文字识别等。项目提供了 OpenAI API 服务、Web UI 示例等部署方式,还包含了视觉处理工具,有助于开发者快速集成和使用,提升工作效率。
HunyuanVideo 是一个可基于文本生成高质量图像和视频的项目。
HunyuanVideo 是一个专注于文本到图像及视频生成的项目。它具备强大的视频生成能力,支持多种分辨率和视频长度选择,能根据用户输入的文本生成逼真的图像和视频。使用先进的技术架构和算法,可灵活调整生成参数,满足不同场景的需求,是文本生成图像视频领域的优质工具。
一个基于 Gradio 构建的 WebUI,支持与浏览器智能体进行便捷交互。
WebUI for Browser Use 是一个强大的项目,它集成了多种大型语言模型,支持自定义浏览器使用,具备持久化浏览器会话等功能。用户可以通过简洁友好的界面轻松控制浏览器智能体完成各类任务,无论是数据提取、网页导航还是表单填写等操作都能高效实现,有利于提高工作效率和获取信息的便捷性。该项目适合开发者、研究人员以及需要自动化浏览器操作的人群使用,在 SEO 优化方面,其关键词涵盖浏览器使用、WebUI、大型语言模型集成等,有助于提高网页在搜索引擎中的曝光度。
基于 ESP32 的小智 AI 开发项目,支持多种网络连接与协议,实现语音交互等功能。
xiaozhi-esp32 是一个极具创新性的基于 ESP32 的开发项目,专注于人工智能语音交互领域。项目涵盖了丰富的功能,如网络连接、OTA 升级、设备激活等,同时支持多种语言。无论是开发爱好者还是专业开发者,都能借助该项目快速搭建起高效的 AI 语音交互系统,为智能设备开发提供强大助力。
一个用于 OCR 的项目,支持多种模型和服务器进行 PDF 到 Markdown 的转换,并提供测试和报告功能。
olmocr 是一个专注于光学字符识别(OCR)的 Python 项目,由 Allen Institute for Artificial Intelligence 开发。它支持多种模型和服务器,如 vllm、sglang、OpenAI 等,可将 PDF 文件的页面转换为 Markdown 格式。项目还提供了测试框架和 HTML 报告生成功能,方便用户对 OCR 结果进行评估和分析。适用于科研、文档处理等领域,有助于提高工作效率和准确性。
飞书多维表格 ×DeepSeek R1 满血版
飞书多维表格联合 DeepSeek R1 模型,提供 AI 自动化解决方案,支持批量写作、数据分析、跨模态处理等功能,适用于电商、短视频、影视创作等场景,提升企业生产力与创作效率。关键词:飞书多维表格、DeepSeek R1、AI 自动化、批量处理、企业协同工具。
最新AI工具、AI资讯
独家AI资源、AI项目落地
微信扫一扫关注公众号