Haozhi Qi

CV & Google Scholar & GitHub & Twitter
I am a Member of Technical Staff at Amazon Frontier AI & Robotics (FAR). My research interests lie broadly in robotics and AI. I develop novel algorithms and robust systems towards the long-term goal of building general-purpose robots that operate in the physical world.
I received my Ph.D. from UC Berkeley, where I was advised by Prof. Yi Ma and Prof. Jitendra Malik. My thesis focused on Multisensory Dexterity for Robotics. I received my Bachelor's degree from Hong Kong University of Science and Technology (HKUST), where I was advised by Prof. Chi-Keung Tang. I also spent two years at Microsoft Research Asia (MSRA), working on computer vision with Dr. Jifeng Dai and Dr. Yichen Wei.

Recent News

Workshops
I co-organize workshops to bring together the dexterous manipulation and tactile sensing community:

Papers (*Equal Technical Contribution, †Equal Advising)

ForceBand: Learning Forceful Manipulation with sEMG
, , , , , , , , †, Haozhi Qi†, †
Tech Report, Jun 2026
Paper Code Website Summary
SPIDER: Scalable Physics-Informed Dexterous Retargeting
, , Haozhi Qi, , , , †, †, †, †
International Conference on Intelligent Robots and Systems (IROS), 2026
Paper Code Website Summary
OSMO: Open-Source Tactile Glove for Human-to-Robot Skill Transfer
, Haozhi Qi*, *, , , , , , ,
Robotics and Automation Letters (RA-L), accepted in 2026
Paper Code Website Summary
HandelBot: Real-World Piano Playing via Fast Adaptation of Dexterous Robot Policies
, Haozhi Qi,
Tech Report, Mar 2026
Paper Code Website Summary
Learning Dexterous Manipulation Skills from Imperfect Simulations
*, *, *, , , †, Haozhi Qi†
International Conference on Robotics and Automation (ICRA), 2026
Paper Code Website Summary
Dexterity from Smart Lenses: Multi-Fingered Robot Manipulation with In-the-Wild Human Demonstrations
, Haozhi Qi, , , , , , , , , , ,
International Conference on Robotics and Automation (ICRA), 2026
Paper Code Website Summary
Coordinated Humanoid Manipulation with Choice Policies
Haozhi Qi*, *, , , , †, †
Tech Report, Dec 2025
Paper Code Website
From Simple to Complex Skills: The Case of In-Hand Object Reorientation
Haozhi Qi, , , , ,
International Conference on Robotics and Automation (ICRA), 2025
Paper Website
Learning Visuotactile Skills with Two Multifingered Hands
, *, *, Haozhi Qi*, , ,
International Conference on Robotics and Automation (ICRA), 2025
Paper Code Website Summary
Learning In-Hand Translation Using Tactile Skin With Shear and Normal Force Sensing
, Haozhi Qi, , , ,
International Conference on Robotics and Automation (ICRA), 2025
Paper Code Website Summary
Hand-Object Interaction Pretraining from Videos
*, *, , , Haozhi Qi, ,
International Conference on Robotics and Automation (ICRA), 2025
Paper Code Website Summary
Lessons from Learning to Spin “Pens”
*, *, *, Haozhi Qi*, , ,
Conference on Robot Learning (CoRL), 2024
Paper Code Website Summary
Twisting Lids Off with Two Hands
*, *, Haozhi Qi, ,
Conference on Robot Learning (CoRL), 2024
Paper Code Website Video Summary
Neural feels with Neural Fields: Visuo-tactile Perception for In-Hand Manipulation
, Haozhi Qi, , , , , , , , , ,
Science Robotics, Nov 2024 (Journal Cover)
Paper Code Website Video Summary
Digitizing Touch with an Artificial Multimodal Fingertip
, , Ali Sengül, Victoria Rose Most, Nolan Black, Kevin Sawyer, Romeo Mercado, Haozhi Qi, Alexander Sohn, Byron Taylor, Norb Tydingco, Gregg Kammerer, Dave Stroud, Jake Khatha, Kurt Jenkins, Kyle Most, Neal Stein, Ricardo Chavira, Thomas Craven-Bartle, Eric Sanchez, Yitian Ding, ,
Tech Report, Nov 2024
Paper Code Website
General In-Hand Object Rotation with Vision and Touch
Haozhi Qi, , , , , ,
Conference on Robot Learning (CoRL), 2023
Paper Website Video Summary
Perceiving Extrinsic Contacts from Touch Improves Learning Insertion Policies
, , Haozhi Qi, , ,
Tech Report, Sep 2023
Paper Code Video Summary
In-Hand Object Rotation via Rapid Motor Adaptation
Haozhi Qi*, *, , ,
Conference on Robot Learning (CoRL), 2022
Paper Code Website Video Summary
Outstanding Demo Award at NeurIPS 2023 Robot Learning Workshop
Coupling Vision and Proprioception for Navigation of Legged Robots
*, *, , Haozhi Qi, ,
Computer Vision and Pattern Recognition (CVPR), 2022
Paper Code Website Video Summary
ReduNet: A White-box Deep Network from the Principle of Maximizing Rate Reduction
*, *, *, Haozhi Qi, ,
Journal of Machine Learning Research (JMLR), accepted in 2022
Paper Code Summary
Ours
GT
Learning Long-term Visual Dynamics with Region Proposal Interaction Networks
Haozhi Qi, , , ,
International Conference on Learning Representations (ICLR), 2021
Paper Code Website Summary
Deep Isometric Learning for Visual Recognition
Haozhi Qi, , , ,
International Conference on Machine Learning (ICML), 2020
Paper Code Website Video Summary
NeurVPS: Neural Vanishing Point Scanning via Conic Convolution
, Haozhi Qi, ,
Neural Information Processing System (NIPS), 2019
Paper Code
Learning to Reconstruct 3D Manhattan Wireframes from a Single Image
, Haozhi Qi, , , , ,
International Conference on Computer Vision (ICCV), 2019 (Oral, 4.3% acceptance rate)
Paper Code
End-to-End Wireframe Parsing
, Haozhi Qi,
International Conference on Computer Vision (ICCV), 2019
Paper Code
Deformable Convolutional Networks
*, Haozhi Qi*, *, *, *, ,
International Conference on Computer Vision (ICCV), 2017 (Oral, 2.1% acceptance rate)
Paper Code
Fully Convolutional Instance-aware Semantic Segmentation
*, Haozhi Qi*, , ,
Computer Vision and Pattern Recognition (CVPR), 2017 (Spotlight, 8.0% acceptance rate)
Paper Code
I highly value open-source and reproducible research. I aim to open-source all of my project code; the current progress is 25 / 27 (92.6%). In addition, my CV template is open-sourced here, and its design is inspired by this. This website is also open-sourced here. Its design is inspired by this and this, with a few adapted elements.
★ ★ ★ ★ ★