Jinpeng Li
Portrait
M. S. @ WHU (2024-Now)
I'm a Master Student at LIESMARS, Wuhan University, under the supervision of Prof. Zhen Dong and Prof. Bisheng Yang, closely working with Assistant Prof. Yuan Liu and Research Assistant Prof. Haiping Wang from HKUST. Before that, I received my B.S. degree from SGES of Sun Yat-sen University. My research focuses on 3D Agent Systems and Explorable World Models.
Experience
Tencent HunYuan

Tencent HunYuan

Research Intern in HunYuan3D Mar. 2026 - Now, Shanghai

WuHan University

WuHan University

Master Student in LIEMSARS Sep. 2024 - Now, Wuhan

Sun Yat-sen University

Sun Yat-sen University

Undergraduate Student in SGES Sep. 2020 - Jul. 2024, Zhuhai

GitHub
News
  • 2026/08: WorldClaw(HunYuan3D Technical Report) is available!
  • 2026/01: SCoT, million-scale CoT dataset for 3D-VLMs, is accepted by ICLR (CCF-A).
  • 2026/01: City-BIS is accepted by JAG (IF: 8.2).
  • 2025/02: CityAnchor, LLM-as-Agent for object grounding, is accepted by ICLR (CCF-A)
  • 2025/01: IPCE-Net is accepted by JAG (IF: 8.2).
  • 2024/03: RoadCorrector is accepted by IEEE TGRS (CCF-B, IF: 9.4).
Honors & Awards
  • Tencent Project UP（青云计划） 2026
  • National Scholarship for Postgraduates (Top 3%) 2023
  • First-Class Scholarship 2023
  • First Prize of National Mathematical Modeling Competition 2023
  • Outstanding Paper At the Geographical Information Science Conference (Top 10) 2023
  • Second-Class Scholarship 2021 & 2022
Selected Publications (view all )
WorldClaw':' Agentic Open-World Generation at Scale

Chunchao Guo, Jinpeng Li, Yang Li, Zilong Huang

HunYuan3D Technical Report (Core Contributor) 2026

We present WorldClaw, an agentic pipeline for fully automatic, high-quality open-world 3D scene generation from text descriptions.

[Paper] [Project Page] [Twitter]

WorldClaw':' Agentic Open-World Generation at Scale

Chunchao Guo, Jinpeng Li, Yang Li, Zilong Huang

HunYuan3D Technical Report (Core Contributor) 2026

We present WorldClaw, an agentic pipeline for fully automatic, high-quality open-world 3D scene generation from text descriptions.

[Paper] [Project Page] [Twitter]

SCoT':' Teaching 3D-LLMs to Think Spatially with Million-scale CoT Annotations
SCoT':' Teaching 3D-LLMs to Think Spatially with Million-scale CoT Annotations

Jinpeng Li, Haiping Wang, Jiabin Chen, Yuan Liu, Zhen Dong†, Bisheng Yang

International Conference on Learning Representations (ICLR) 2026

We present a million-scale 3D visual-language dataset with CoT annotations that unifies perception, analysis, and planning tasks to advance interpretable 3D intelligence.

[Paper] [Code]

SCoT':' Teaching 3D-LLMs to Think Spatially with Million-scale CoT Annotations

Jinpeng Li, Haiping Wang, Jiabin Chen, Yuan Liu, Zhen Dong†, Bisheng Yang

International Conference on Learning Representations (ICLR) 2026

We present a million-scale 3D visual-language dataset with CoT annotations that unifies perception, analysis, and planning tasks to advance interpretable 3D intelligence.

[Paper] [Code]

CityAnchor':' City-scale 3D Visual Grounding with Multi-modality LLMs
CityAnchor':' City-scale 3D Visual Grounding with Multi-modality LLMs

Jinpeng Li, Haiping Wang, Jiabin Chen, Yuan Liu†, Zhiyang Dou, Yuexin Ma, Sibei Yang, Yuan Li, Wenping Wang, Zhen Dong, Bisheng Yang†

International Conference on Learning Representations (ICLR) 2025

We present a two-stage (coarse-to-fine) 3D visual grounding system by tuning Large Vision Language Model (LVLM) to accurately find targets in city-scale point clouds from text descriptions.

[Paper] [Code]

CityAnchor':' City-scale 3D Visual Grounding with Multi-modality LLMs

Jinpeng Li, Haiping Wang, Jiabin Chen, Yuan Liu†, Zhiyang Dou, Yuexin Ma, Sibei Yang, Yuan Li, Wenping Wang, Zhen Dong, Bisheng Yang†

International Conference on Learning Representations (ICLR) 2025

We present a two-stage (coarse-to-fine) 3D visual grounding system by tuning Large Vision Language Model (LVLM) to accurately find targets in city-scale point clouds from text descriptions.

[Paper] [Code]

3DCity-LLM':' Empowering Multi-modality Large Language Models for 3D City-scale Perception and Understanding
3DCity-LLM':' Empowering Multi-modality Large Language Models for 3D City-scale Perception and Understanding

Yiping Chen*†, Jinpeng Li*, Wenyu Ke, Yang Luo, Jie Ouyang, Zhongjie He, Li Liu, Hongchao Fan, Hao Wu (* equal contribution)

Arxiv 2026

We propose 3DCity-LLM for 3D city-scale vision-language perception and understanding, with 3DCity-LLM-1.2M dataset that comprising approximately 1.2 million high-quality samples to facilitate large-scale training.

[Paper] [Code]

3DCity-LLM':' Empowering Multi-modality Large Language Models for 3D City-scale Perception and Understanding

Yiping Chen*†, Jinpeng Li*, Wenyu Ke, Yang Luo, Jie Ouyang, Zhongjie He, Li Liu, Hongchao Fan, Hao Wu (* equal contribution)

Arxiv 2026

We propose 3DCity-LLM for 3D city-scale vision-language perception and understanding, with 3DCity-LLM-1.2M dataset that comprising approximately 1.2 million high-quality samples to facilitate large-scale training.

[Paper] [Code]

SpatialLLM':' From Multi-modality Data to Urban Spatial Intelligence
SpatialLLM':' From Multi-modality Data to Urban Spatial Intelligence

Jiabin Chen*, Haiping Wang*, Jinpeng Li, Yuan Liu†, Zhen Dong†, Bisheng Yang (* equal contribution)

International Journal of Applied Earth Observation and Geoinformation (IF:8.2) 2026

Structured descriptions of raw spatial data equip LLM with zero-shot execution of advanced spatial intelligence tasks, including urban planning, ecological analysis, traffic management, etc.. Multi-field knowledge, context length, and reasoning ability are key factors influencing LLM performances in urban analysis.

[Paper] [Code]

SpatialLLM':' From Multi-modality Data to Urban Spatial Intelligence

Jiabin Chen*, Haiping Wang*, Jinpeng Li, Yuan Liu†, Zhen Dong†, Bisheng Yang (* equal contribution)

International Journal of Applied Earth Observation and Geoinformation (IF:8.2) 2026

Structured descriptions of raw spatial data equip LLM with zero-shot execution of advanced spatial intelligence tasks, including urban planning, ecological analysis, traffic management, etc.. Multi-field knowledge, context length, and reasoning ability are key factors influencing LLM performances in urban analysis.

[Paper] [Code]

RoadCorrector: A Structure-Aware Road Extraction Method for Road Connectivity and Topology Correction
RoadCorrector: A Structure-Aware Road Extraction Method for Road Connectivity and Topology Correction

Jinpeng Li, Jun He, Weijia Li, Jiabin Chen, Jinhua Yu

IEEE Transactions on Geoscience and Remote Sensing (T-GRS, CCF-B, IF:9.6) 2024

We propose RoadCorrector to enhance the road integrity and connectivity by adding structure-related assistance branches and two correction modules

[Paper]

RoadCorrector: A Structure-Aware Road Extraction Method for Road Connectivity and Topology Correction

Jinpeng Li, Jun He, Weijia Li, Jiabin Chen, Jinhua Yu

IEEE Transactions on Geoscience and Remote Sensing (T-GRS, CCF-B, IF:9.6) 2024

We propose RoadCorrector to enhance the road integrity and connectivity by adding structure-related assistance branches and two correction modules

[Paper]

IPCE-Net':' Image-point cloud embedding network for simultaneous image-based farmland instance extraction and point cloud-based semantic segmentation
IPCE-Net':' Image-point cloud embedding network for simultaneous image-based farmland instance extraction and point cloud-based semantic segmentation

Jinpeng Li, Yuan Li†, Shuhang Zhang, Yiping Chen

International Journal of Applied Earth Observation and Geoinformation (IF:8.6) 2025

We propose an end-to-end bimodal network IPCE-Net for simultaneous image and point cloud segmentation.

[Paper]

IPCE-Net':' Image-point cloud embedding network for simultaneous image-based farmland instance extraction and point cloud-based semantic segmentation

Jinpeng Li, Yuan Li†, Shuhang Zhang, Yiping Chen

International Journal of Applied Earth Observation and Geoinformation (IF:8.6) 2025

We propose an end-to-end bimodal network IPCE-Net for simultaneous image and point cloud segmentation.

[Paper]

City-BIS':' City-scale building instance segmentation from LiDAR point clouds via structure-aware method
City-BIS':' City-scale building instance segmentation from LiDAR point clouds via structure-aware method

Jinpeng Li, Yuan Li, Yiping Chen†, Hongchao Fan, Ruisheng Wang

International Journal of Applied Earth Observation and Geoinformation (IF:8.2) 2026

We present a novel pipeline to achieve both efficient building semantic segmentation and robust building instance segmentation in large-scale scene.

[Paper] [Code]

City-BIS':' City-scale building instance segmentation from LiDAR point clouds via structure-aware method

Jinpeng Li, Yuan Li, Yiping Chen†, Hongchao Fan, Ruisheng Wang

International Journal of Applied Earth Observation and Geoinformation (IF:8.2) 2026

We present a novel pipeline to achieve both efficient building semantic segmentation and robust building instance segmentation in large-scale scene.

[Paper] [Code]

All publications