返回团队成员列表
连文昭

连文昭 (Wenzhao Lian)

教授
lianwenzhao@sjtu.edu.cn GitHub

个人简介

教授，国家级青年人才，主要研究领域包括智能机器人、灵巧操作和机器学习。于2015年在美国杜克大学获得博士学位，曾任Google X资深研究科学家和Figure AI技术总监。此前研发了基于云边协同和人工智能的机器人操作系统平台，搭建了软硬一体的类人形机器人，落地了多种形态的工业级产品，申请国际专利十余项。在机器人和机器学习领域的顶级学术会议与期刊(ICRA/RSS/IROS/ICML/Neurips等)发表学术论文数十篇，包括实现了通过单一样本示教习得可泛化的复杂操作技能，操作精度达到亚毫米，提名RSS最佳论文。

研究方向

智能机器人 (Intelligent Robotics)

灵巧操作 (Dexterous Manipulation)

机器学习 (Machine Learning)

研究项目

1.基于云边协同和人工智能的机器人编程软件

在机器人操作系统层面，团队开发了全球首个面向工业机器人的基于云边协同与AI驱动的跨平台编程系统，实现了用户无需具备专业知识即可进行机器人任务编排。该系统显著提升了工业机器人在复杂装配场景（如家具、电子元器件）中的灵活性与可部署性，技术已获得8项国际专利，成果成功转化为Intrinsic公司首发产品并在Google X考核中获得最高评分，交付全球多家车企与消费电子厂商使用。

2.通过单一样本示教习得可泛化的复杂操作技能

在机器人技能学习方面，团队实现了单一样本示教即可泛化复杂操作技能的突破。该技术通过创新的视觉感知与仿真学习方法，摆脱了对大量真实数据的依赖，仅需单一示范即可实现亚毫米级精度的复杂操作。团队提出的基于视觉的物体抓取与操作方法，能够在仿真环境中学习技能并直接迁移至真实机器人，解决了传统方法在非结构化环境中的适应性问题。该成果已在ICRA、RSS等机器人领域顶级会议发表，并在Google X实现技术落地，为农业场景中的精细操作提供了核心技术支撑。

3.创新型灵巧操作技能学习框架

在系统架构层面，团队构建了高性能的端到端灵巧操作框架。该框架集成了意图识别、多模态感知、实时反馈控制与故障恢复机制，突破了传统方法在复杂操作任务中的性能瓶颈。其中多模态感知模块融合视觉、触觉等多种传感信息，实时反馈控制模块确保操作的精确执行，故障恢复机制则提升了系统的鲁棒性。该框架获得IROS评审委员会的高度评价，被认为是对实际部署场景具有深刻洞察的系统性工作。

发表论文

  • 2026 arXiv

    BridgeACT: Bridging Human Demonstrations to Robot Actions via Unified Tool-Target Affordances

    Yifan Han, Jianxiang Liu, Haoyu Zhang, Yuqi Gu, Yunhan Guo, Wenzhao Lian

    📄 arXiv
  • 2026 arXiv

    DexHiL: A Human-in-the-Loop Framework for Vision-Language-Action Model Post-Training in Dexterous Manipulation

    Yifan Han, Zhongxi Chen, Yuxuan Zhao, Congsheng Xu, Yanming Shao, Yichuan Peng, Yao Mu, Wenzhao Lian

    📄 arXiv
  • 2026 arXiv

    CLASH: Collision Learning via Augmented Sim-to-real Hybridization to Bridge the Reality Gap

    Haotian He, Ning Guo, Siqi Shi, Qipeng Liu, Wenzhao Lian

    📄 arXiv
  • 2026 RSS (在投)

    Act, Sense, Act: Learning Non-Markovian Active Perception Strategies from Large-Scale Egocentric Human Data

    Jialiang Li, Yi Qiao, Yunhan Guo, Changwen Chen, Wenzhao Lian

    📄 arXiv 🌐 项目主页
  • 2026 arXiv

    Viewpoint Matters: Dynamically Optimizing Viewpoints with Masked Autoencoder for Visual Manipulation

    Pengfei Yi, Yifan Han, Junyan Li, Litao Liu, Wenzhao Lian

    📄 arXiv
  • 2026 arXiv

    EmboCoach-Bench: Benchmarking AI Agents on Developing Embodied Robots

    Zixing Lei, Genjia Liu, Yuanshuo Zhang, Qipeng Liu, Chuan Wen, Shanghang Zhang, Wenzhao Lian, Siheng Chen

    📄 arXiv
  • 2026 arXiv

    FSAG: Enhancing Human-to-Dexterous-Hand Finger-Specific Affordance Grounding via Diffusion Models

    Yifan Han, Yichuan Peng, Pengfei Yi, Junyan Li, Hanqing Wang, Gaojing Zhang, Qi Peng Liu, Wenzhao Lian

    📄 arXiv
  • 2026 ICRA

    SAGE: Scene Graph-Aware Guidance and Execution for Long-Horizon Manipulation Tasks

    Jialiang Li, Wenzheng Wu, Gaojing Zhang, Yifan Han, Wenzhao Lian

    📄 arXiv
  • 2026 AAAI

    FoAM: Foresight-Augmented Multi-Task Imitation Policy for Robotic Manipulation

    Litao Liu, Wentao Wang, Yifan Han, Zhuoli Xie, Pengfei Yi, Junyan Li, Yi Qin, Wenzhao Lian

    📄 arXiv
  • 2025 arXiv

    Look, Zoom, Understand: The Robotic Eyeball for Embodied Perception

    Jiashu Yang, Yifan Han, Yucheng Xie, Ning Guo, Wenzhao Lian

    📄 arXiv
  • 2025 IROS

    Splatter Joint: 3D Gaussian Splatting for Articulated Objects

    Junyan Li, Yifan Han, Pengfei Yi, Wenzhao Lian

    📄 arXiv
  • 2024 IROS

    Stick Roller: Precise In-Hand Stick Rolling with a Sample-Efficient Tactile Model

    Yipai Du, Pokuang Zhou, Michael Yu Wang, Wenzhao Lian, Yu She

  • 2023 ICRA

    Allowing Safe Contact in Robotic Goal-Reaching: Planning and Tracking in Operational and Null Spaces

    Xinghao Zhu, Wenzhao Lian, Bodi Yuan, Daniel Freeman, Masayoshi Tomizuka

  • 2023 ICRA

    Cable Routing and Assembly using Tactile-driven Motion Primitives

    Achu Wilson, Helen Jiang, Wenzhao Lian, Wenzhen Yuan

  • 2023 ICRA

    Zero-Shot Policy Transfer with Disentangled Task Representation of Meta-Reinforcement Learning

    Zheng Wu, Yichen Xie, Wenzhao Lian, Changhao Wang, Yanjiang Guo, Jianyu Chen, Stefan Schaal, Masayoshi Tomizuka

  • 2023 IFAC

    Prim-LAfD: A Framework to Learn and Adapt Primitive-Based Skills from Demonstrations for Insertion Tasks

    Zheng Wu, Wenzhao Lian, Changhao Wang, Mengxi Li, Stefan Schaal, Masayoshi Tomizuka

  • 2022 RSS (Best Paper Award Finalist)

    You Only Demonstrate Once: Category-Level Manipulation from Single Visual Demonstration

    Bowen Wen, Wenzhao Lian, Kostas Bekris, Stefan Schaal

  • 2022 ICRA

    CaTGrasp: Learning Category-Level Task-Relevant Grasping in Clutter from Simulation

    Bowen Wen, Wenzhao Lian, Kostas Bekris, Stefan Schaal

  • 2022 ICRA

    Symbolic State Estimation with Predicates for Contact-Rich Manipulation Tasks

    Toki Migimatsu, Wenzhao Lian, Jeannette Bohg, Stefan Schaal

  • 2021 IROS

    Benchmarking Off-The-Shelf Solutions to Robotic Assembly Tasks

    Wenzhao Lian, Tim Kelch, Dirk Holz, Adam Norton, Stefan Schaal

  • 2021 RA-L

    Robotic Cable Routing with Spatial Representation

    Shiyu Jin, Wenzhao Lian, Changhao Wang, Masayoshi Tomizuka, Stefan Schaal

  • 2021 RSS

    Robust Multi-Modal Policies for Industrial Assembly via Reinforcement Learning and Demonstrations: A Large-Scale Study

    Jianlan Luo, Oleg Sushkov, Rugile Pevceviciute, Wenzhao Lian, Chang Su, Mel Vecerik, Ning Ye, Stefan Schaal, Jon Scholz

  • 2021 ICRA

    Learning Dense Rewards for Contact-Rich Manipulation Tasks

    Zheng Wu, Wenzhao Lian, Vaibhav Unhelkar, Masayoshi Tomizuka, Stefan Schaal

  • 2021 ICRA

    Interpreting Contact Interactions to Overcome Failure in Robot Assembly Tasks

    Peter A. Zachares, Michelle A. Lee, Wenzhao Lian, Jeannette Bohg

  • 2017 KDD

    Convex factorization machine for toxicogenomics prediction

    Makoto Yamada, Wenzhao Lian, Amit Goyal, Jianhui Chen, Kishan Wimalawarne, Suleiman A Khan, Samuel Kaski, Hiroshi Mamitsuka, Yi Chang

  • 2015 ICML (Appendix)

    A Multitask Point Process Predictive Model

    Wenzhao Lian, Ricardo Henao, Vinayak Rao, Joseph Locus, Lawrence Carin

  • 2015 AAAI (Appendix)

    Integrating Features and Similarities: Flexible Models for Heterogeneous Multiview Data

    Wenzhao Lian, Piyush Rai, Esther Salazar, Lawrence Carin

  • 2015 COSI

    Task-driven Adaptive Sensing on Quadrupole Mass Filter Systems for Classification

    Evan X. Chen, Wenzhao Lian, Lawrence Carin, David Brady

  • 2015 Signal Process.

    Multivariate Time-Series Analysis and Diffusion Maps

    Wenzhao Lian, Ronen Talmen, Hitten Zaveri, Lawrence Carin, Ronald Coifman

  • 2014 NIPS (Appendix)

    Analysis of Brain States from Multi-Region LFP Time-Series

    Kyle Ulrich, David Carlson, Wenzhao Lian, Jana Borg, Kafui Dzirasa, Lawrence Carin

  • 2014 Submitted

    A Bayesian Framework for Multi-Modality Analysis of Mental Health

    Esther Salazar, Yuliya Nikolova, Wenzhao Lian, Piyush Rai, Adrienne L. Romer, Ahmad R. Hariri, Lawrence Carin

  • 2014 NIPS Workshop

    Bayesian Multitask Distance Metric Learning

    Piyush Rai, Wenzhao Lian, Lawrence Carin

  • 2014 NIPS Workshop

    Bayesian Multiview Factor Modeling for Integrating and Analyzing Heterogeneous Clinical Data

    Wenzhao Lian, Piyush Rai, Esther Salazar, Lawrence Carin

  • 2014 ICML (Appendix)

    Modeling Correlated Arrival Events with Latent Semi-Markov Processes

    Wenzhao Lian, Vinayak Rao, Brian Eriksson, Lawrence Carin

  • 2013 TBME

    Sorting Electrophysiological Data via Dictionary Learning & Mixture Modeling

    David E. Carlson, Joshua T. Vogelstein, Qisong Wu, Wenzhao Lian, Mingyuan Zhou, Colin R. Stoetzner, Daryl Kipke, Douglas Weber, David B. Dunson, Lawrence Carin