About Me

I received my M.S. from the Department of Computer Science at Tsinghua University, advised by Prof. Jun Zhu. In 2022, I obtained my B.S. in the School of Mathematical Sciences at Peking University. I am currently a Researcher at Tencent Hunyuan.

My research interests lie in 3D AIGC (e.g., artist mesh generation, part generation, 3D editing), unified multimodal models (UMM), and embodied intelligence.

Publications

* equal contribution † project leader ‡ core contributor
Topic
Authorship
Hunyuan3D-Buffalo 1.0
Hunyuan3D-Buffalo 1.0: A Unified Multimodal Model for Scalable 3D Generation, Understanding, and Editing
Tech Report 2026
Junliang Ye‡, Kenkun Liu‡, Guocun Wang‡, Yang Li‡†, Yansong Qu‡, Chunshi Wang‡, et al. (Tencent Hunyuan)
A unified 3D multimodal framework for 3D understanding, generation, and editing.
Paper Project
PolyFlow
PolyFlow: Continuous Topology Embedding Flow Matching for Artist-style Mesh Generation
SIGGRAPH ASIA 2026
Chunshi Wang*, Haohan Weng*, Junliang Ye*, Biwen Lei, Yang Li, Zibo Zhao, et al.
A Transformer-based flow-matching framework for parallel artist-style mesh generation with continuous topology embedding, achieving faster inference and precise vertex-count control.
Paper
GEM
GEM: Generative Supervision Helps Embodied Intelligence
ECCV 2026
Ruowen Zhao, Bangguo Li, Zuyan Liu, Yinan Liang, Junliang Ye, et al.
A generative-supervised embodied VLM that enhances semantic reasoning and physical grounding by integrating depth map generation into VLM pre-training.
Paper Code Project stars
PhysForge
PhysForge: Generating Physics-Grounded 3D Assets for Interactive Virtual World
ICML 2026
Yunhan Yang*, Chunshi Wang*, Junliang Ye*, Yang Li, Zanxin Chen, Zehuan Huang, et al.
A decoupled two-stage framework supported by PhysDB, a large-scale dataset of 150,000 assets with four-tier physical annotations.
Paper Code Project stars
UniVerse3D
UniVerse3D: Emerging Properties of Unified Multimodal Models in 3D Understanding and Generation
CVPR-Findings 2026
Junliang Ye*, Zehuan Huang*, Yansong Qu*, Chunshi Wang, Yunhan Yang, Yang Li, et al.
We propose UniVerse3D, the first 3D Unified multimodal models.
Paper
ReconX
ReconX: Reconstruct Any Scene from Sparse Views with Video Diffusion Model
TIP 2026
Fangfu Liu*, Wenqiang Sun*, Hanyang Wang*, Yikai Wang, Haowen Sun, Junliang Ye, et al.
Paper Code Project stars
Nano3D
Nano3D: A Training-Free Approach for Efficient 3D Editing Without Masks
ICLR 2026
Junliang Ye*, Shenghao Xie*, Ruowen Zhao, Zhengyi Wang, Hongyu Yan, Wenqiang Zu, et al.
A training-free framework for precise and coherent 3D object editing without masks.
Paper Code Project stars
Part-X-MLLM
Part-X-MLLM: Part-aware 3D Multimodal Large Language Model
ICLR 2026
Chunshi Wang*, Junliang Ye†*, Yunhan Yang*, Yang Li, Zizhuo Lin, Jun Zhu, et al.
A native 3D multimodal large language model that unifies diverse 3D tasks via structured, executable grammar.
Paper Code Project stars
DeepMesh-v2
DeepMesh-v2: Auto-Regressive Artist-Mesh Creation With Reinforcement Learning
arXiv 2025
Junliang Ye*, Ruowen Zhao*, Zhengyi Wang*, Yikai Wang, Jun Zhu
Code stars
DreamReward-X
DreamReward-X: Boosting High-Quality 3D Generation with Human Preference Alignment
TPAMI 2025
Fangfu Liu, Junliang Ye, Hanyang Wang, Zhengyi Wang, Jun Zhu, Yueqi Duan
Project
ShapeLLM-Omni
ShapeLLM-Omni: A Native Multimodal LLM for 3D Generation and Understanding
NeurIPS 2025Spotlight
Junliang Ye*, Zhengyi Wang*, Ruowen Zhao*, Shenghao Xie, Jun Zhu
A multimodal large model that integrates 3D generation, understanding, and editing capabilities.
Paper Code Project stars
DeepMesh
DeepMesh: Auto-Regressive Artist-Mesh Creation With Reinforcement Learning
ICCV 2025
Ruowen Zhao*, Junliang Ye*, Zhengyi Wang*, Guangce Liu, Yiwen Chen, Yikai Wang, et al.
Generates meshes with intricate details and precise topology, surpassing SOTA in both precision and quality.
Paper Code Project stars
DreamReward
DreamReward: Aligning Human Preference in Text-to-3D Generation
ECCV 2024
Junliang Ye*, Fangfu Liu*, Qixiu Li, Zhengyi Wang, Yikai Wang, Xinzhou Wang, et al.
A comprehensive framework to learn and improve text-to-3D models from human preference feedback.
Paper Code Project stars
AnimatableDreamer
AnimatableDreamer: Text-Guided Non-rigid 3D Model Generation and Reconstruction
ECCV 2024
Xinzhou Wang, Yikai Wang, Junliang Ye, Fucun Sun, Zhengyi Wang, Ling Wang, et al.
Paper Project

Experience & Education

Work

Researcher · Tencent Hunyuan 2026.08 – Now
Working on 3D Unified Multimodal Models.
Qingyun Research Intern · Tencent Hunyuan 2025.07 – 2026.01
Working on 3D Unified Multimodal Models.
Research Intern · ShengShu Technology 2024.03 – 2025.06
Working on 3D and video generation research.

Education

M.S., Computer Science · Tsinghua University 2023 – 2026
Advisor: Prof. Jun Zhu
B.S., School of Mathematical Sciences · Peking University 2018 – 2022

Awards

Outstanding Graduate of Tsinghua University 2026 (Top 2%)
Outstanding Graduate of Department of Computer Science, Tsinghua University 2026
Chinese Government Scholarship 2023 – 2026
© Junliang Ye · 叶俊良