Hao Chen
Research Professor
Builds vision foundation models, including repurposing generative models for perception tasks (DiCeption, GenPercept) and metric depth estimation (Metric3D).
Hao Chen's lab at Zhejiang works on 3D vision & neural rendering, generative & foundation models and robot perception & SLAM. 54 papers since 2023, 48 at top venues, mostly ICLR and CVPR. Applications close .
At a glance
- papers, 2023–2026
- 14/ year
- at T1 venues
- 89%
- PhD students
- —
- MS students
- —
- alumni listed
- —
Papers by yearfrom the lab homepage, 2023–2026
- 14 papers2026 · T1 13 · collecting
- 14 papers2025 · T1 13
- 17 papers2024 · T1 13
- 9 papers2023 · T1 9
T1 venuesOtherVenues: ICLR, CVPR, ICCV, ICML, AAAI
Research focuspaper titles that use each topic's words
- Generative & foundation models14 papers
- Robot perception & SLAM6 papers
- 3D vision & neural rendering3 papers
- depth3 papers
- depth estimation1 paper
Paper list
From the lab homepage.
202614 papers
- ICML (T1)ACTIVE-o3: Empowering MLLMs with Active Perception via Pure Reinforcement Learning
- —Beyond Hard Masks: Progressive Token Evolution for Diffusion Language Models
- ACL (T1)Efficient Self-Evaluation for Diffusion Language Models via Sequence Regeneration
- CVPR (T1)Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching
- CVPR (T1)Exploring Spatial Intelligence from a Generative Perspective
- IJCV (T1)FreerCustom: Training-Free Multi-Concept Customization for Image and Video Generation: C. Zhao et al
- ICML (T1)GAE: Unleashing Physical Potential of VLM with Generalizable Action Expert
- AAAI (T1)Odyssey: Open-world quadrupeds exploration and manipulation for long-horizon tasks
- CVPR (T1)Preserving Source Video Realism: High-Fidelity Face Swapping for Cinematic Quality
- ICML (T1)SpeedVFI: One-step Diffusion for Efficient Video Frame Interpolation
- CVPR (T1)Stamo: Unsupervised learning of generalizable robot motion from compact state representation
- ICLR (T1)Time is a feature: Exploiting temporal dynamics in diffusion language models
202514 papers
- ICLR (T1)Boltzmann-Aligned Inverse Folding Model as a Predictor of Mutational Effects on Protein-Protein Interactions
- NeurIPS (T1)Diception: A generalist diffusion model for visual perceptual tasks
- TPAMI (T1)Diffusion Models are Efficient Data Generators for Human Mesh Recovery (opens in a new tab)
- ICLR (T1)Framer: Interactive Frame Interpolation
- Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference PapersGenerative Video Matting (opens in a new tab)
- ICLR (T1)MovieDreamer: Hierarchical Generation for Coherent Long Visual Sequences
- NeurIPS (T1)Omni-R1: Reinforcement Learning for Omnimodal Reasoning via Two-System Collaboration
- ICLR (T1)PerturboLLaVA: Reducing Multimodal Hallucinations with Perturbative Visual Training
- ICCV (T1)POMATO: Marrying Pointmap Matching with Temporal Motion for Dynamic 3D Reconstruction
- ICLR (T1)Revisiting Convolution Architecture in the Realm of DNA Foundation Models
- CVPR (T1)SegAgent: Exploring Pixel Understanding Capabilities in MLLMs by Imitating Human Annotator Trajectories
- IJCV (T1)Segment Anything in Context with Vision Foundation Models
202417 papers
- Hao ChenA Simple Image Segmentation Framework via In-Context Examples
- IJCV (T1)AutoStory: Generating Diverse Storytelling Images with Minimal Human Efforts (opens in a new tab)
- ICLR (T1)De Novo Protein Design Using Geometric Vector Field Networks
- AAAI (T1)DiffCalib: Reformulating Monocular Camera Calibration as Diffusion-Based Dense Incident Map Generation
- CVPR (T1)DiverGen: Improving Instance Segmentation by Learning Wider Data Distribution with More Diverse Generative Data
- ICML (T1)Floating Anchor Diffusion Model for Multi-motif Scaffolding
- The European Conference on Computer Vision ECCV , AugFreeCompose: Generic Zero-Shot Image Composition with Diffusion Prior
- CVPR (T1)FreeCustom: Tuning-Free Customized Image Generation for Multi-Concept Composition
- ICML (T1)Generative Active Learning for Long-tailed Instance Segmentation
- Findings of ACLLoRAPrune: Structured Pruning Meets Low-Rank Parameter-Efficient Fine-Tuning (opens in a new tab)
- Hao ChenLSSInst: Improving Geometric Modeling in LSS-Based BEV Perception with Instance Representation
- ICLR (T1)Matcher: Segment Anything with One Shot Using All-Purpose Feature Matching
20239 papers
- IJCV (T1)A Dynamic Feature Interaction Framework for Multi-task Visual Perception (opens in a new tab)
- ICCV (T1)CTVIS: Consistent Training for Online Video Instance Segmentation
- NeurIPS (T1)DatasetDM: Synthesizing Data with Perception Annotations Using Diffusion Models
- ICCV (T1)FrozenRecon: Pose-free 3D Scene Reconstruction with Frozen Depth Models
- CVPR (T1)Learning Conditional Attributes for Compositional Zero-Shot Learning
- CVPR (T1)Learning To Fuse Monocular and Multi-View Cues for Multi-Frame Depth Estimation in Dynamic Scenes
- ICCV (T1)Metric3D: Towards Zero-shot Metric 3D Prediction from A Single Image
- ICLR (T1)Object-Aware Inversion and Reassembly for Image Editing
- ICCV (T1)SegPrompt: Boosting Open-World Segmentation via Category-Level Prompt Learning
Where alumni went
Alumni are not listed on the lab homepage.
Apply to ZhejiangFall 2027
Computer Science PhD (international applicants)
- Round
- Self-funded
- Opens
- —
- GRE
- See program page
- English
- Per ZJU language requirements
- Fee
- RMB 800
Not yet confirmed for this intake.
Similar labsnearest by research summary
- USCMU
Fernando De la Torre
- Embodied & active vision
- CNTsinghua
Song-Chun Zhu
- Vision-language & multimodal
- GBImperial
Krystian Mikolajczyk
- 3D vision & neural rendering
- GBUCL
Sophia Bano
- Recognition, detection & segmentation
- USGeorgia Tech
Zsolt Kira
- Computational imaging & photography
- CNZhejiang
Mingli Song
113 T1
Your notes
Save this lab to keep notes.
Data
Some numbers could not be verified:
- The member list could not be found on the lab site.
- The alumni list could not be found on the lab site.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Not collected
- Alumni
- Not collected
Tiers are venue classifications, not lab rankings.
Zhejiangcloses