Huazhe Xu
Assistant Professor
Works on deep reinforcement learning, robotics, computer vision, and tactile sensing.
Huazhe Xu's lab at Tsinghua works on robot perception & SLAM and reinforcement learning. 58 papers since 2023, 57 at top venues, mostly ICLR and ICML. 11 PhD students and 1 master's student. Applications close .
At a glance
- papers, 2023–2026
- 15/ year
- at T1 venues
- 98%
- PhD students
- 11
- MS students
- 1
- alumni listed
- —
Papers by yearfrom the lab homepage, 2023–2026
- 0 papers2026 · T1 0 · collecting
- 17 papers2025 · T1 17
- 25 papers2024 · T1 24
- 16 papers2023 · T1 16
T1 venuesOtherVenues: ICLR, ICML, ICRA, NeurIPS, CoRL
Research focuspaper titles that use each topic's words
- Reinforcement learning28 papers
- Robot perception & SLAM4 papers
- tactile3 papers
- deep reinforcement2 papers
- tactile sensing0 papers
Paper list
From the lab homepage.
202517 papers
- ICML (T1)A Forget-and-Grow Strategy for Deep Reinforcement Learning Scaling in Continuous Control
- ICRA (T1)Catch It! Learning to Catch in Flight with Mobile Dexterous Hands
- RSS (T1)DemoGen: Synthetic Demonstration Generation for Data-Efficient Visuomotor Policy Learning
- CoRL (T1)DemoSpeedup: Accelerating Visuomotor Policies via Entropy-Guided Demonstration Acceleration
- ICLR (T1)DenseMatcher: Learning 3D Semantic Correspondence for Category-Level Manipulation from a Single Demo
- RSS (T1)DOGlove: Dexterous Manipulation with a Low-Cost Open-Source Haptic Force Feedback Glove
- IROS (T1)DTactive: A Vision-Based Tactile Sensor with Active Surface
- CoRL (T1)FACET: Force-Adaptive Control via Impedance Reference Tracking for Legged Robots
- ICRA (T1)Fine-Tuning Hard-to-Simulate Objectives for Quadruped Locomotion: A Case Study on Total Power Saving
- ICLR (T1)Looking Backward: Retrospective Backward Synthesis for Goal-Conditioned GFlowNets
- ICML (T1)MENTOR: Mixture-of-Experts Network with Task-Oriented Perturbation for Visual Reinforcement Learning
- RSS (T1)Morpheus: A Neural-driven Animatronic Face with Hybrid Actuation and Diverse Emotion Control
202425 papers
- RSS (T1)3D Diffusion Policy
- ICRA (T1)A Wearable Robotic Hand for Hand-over-Hand Imitation Learning
- ICML (T1)ACE: Off-Policy Actor-Critic with Causality-Aware Entropy Regularization
- ICRA (T1)ArrayBot: Reinforcement Learning for Generalizable Distributed Manipulation through Touch
- ICLR (T1)COPlanner: Plan to Roll Out Conservatively but to Explore Optimistically for Model-Based RL
- ICRA (T1)DeformNet: Latent Space Modeling and Dynamics Prediction for Deformable Object Manipulation
- ECCV (T1)Diffusion Reward: Learning Rewards via Conditional Video Diffusion
- ICLR (T1)DrM: Mastering Visual Reinforcement Learning through Dormant Ratio Minimization
- ICLR (T1)GenSim: Generating Robotic Simulation Tasks via Large Language Models
- CoRL (T1)GenSim2: Scaling Robotic Data Generation with Multi-modal and Reasoning LLMs
- NeurIPS (T1)Key-Grid: Unsupervised 3D Keypoints Detection using Grid Heatmap Features
- CoRL (T1)Learning to Manipulate Anywhere: A Visual Generalizable Framework For Reinforcement Learning
202316 papers
- NeurIPS (T1)Can Pre-Trained Text-to-Image Models Generate Visual Goals for Reinforcement Learning?
- NeurIPS (T1)CEIL: Generalized Contextual Imitation Learning
- ICLR (T1)Decision Transformer under Random Frame Dropping
- ICRA (T1)DTact: A Vision-Based Tactile Sensor that Measures High-Resolution 3D Geometry Directly from Darkness
- ICRA (T1)Efficient Bimanual Handover and Rearrangement via Symmetry-Aware Actor-Critic Learning
- ICRA (T1)Extraneousness-Aware Imitation Learning
- NeurIPS (T1)H-InDex: Visual Reinforcement Learning with Hand-Informed Representations for Dexterous Manipulation
- ICLR (T1)Is Model Ensemble Necessary? Model-based RL via a Single Model with Lipschitz Regularized Value Function
- NeurIPS (T1)MoVie: Visual Model-Based Policy Adaptation for View Generalization
- ICML (T1)On Pre-Training for Visuo-Motor Control: Revisiting a Learning-from-Scratch Baseline
- NeurIPS (T1)RL-ViGen: A Reinforcement Learning Benchmark for Visual Generalization
- CoRL (T1)RoboCook: Long-Horizon Elasto-Plastic Object Manipulation with Diverse Tools
Where alumni went
Alumni are not listed on the lab homepage.
Apply to TsinghuaFall 2027
Computer Science PhD (international applicants)
- Round
- Main
- Opens
- GRE
- See program page
- English
- Per school catalog
- Fee
- RMB 800
Not yet confirmed for this intake.
Similar labsnearest by research summary
- GBUCL
Yasemin Bekiroglu
- Manipulation & robot learning
- CNPeking
Hong Liu
- Robot perception & SLAM
- USUT Austin
Sandeep Chinchali
- Planning, control & multi-robot systems
- USCMU
Oliver Kroemer
- Manipulation & robot learning
- USMichigan
Nima Fazeli
- Manipulation & robot learning
- USNYU
Chen Feng
- Robot perception & SLAM
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
Tsinghuacloses