yw.

RESEARCHER · BUILDER · EXPLORER

Yifu Wang王一夫 / Ian

Perceive the world.
Create new possibilities.

Founding Member & COO at Vertex Lab | V2Fun.ai
Generative worlds, digital humans and spatial intelligence.

Google Scholar ↗
GENERATIVE AI3D & VIDEOSPATIAL PERCEPTION
Yifu Wang

01 / A LITTLE ABOUT ME

A vision researcher.
A builder at heart.

My background ↗

I study how machines perceive, reconstruct and generate the visual world. My research spans generative AI for 3D scenes, consistent and controllable video generation, and expressive digital avatars, alongside geometric vision, visual-inertial SLAM and event-based perception. Across these areas, I explore how spatial understanding can help machines navigate real environments and create coherent, editable virtual worlds.

I am a Founding Member and COO at Vertex Lab | V2Fun.ai. Previously, I worked as a Senior Researcher at Tencent XR Vision Labs and as a Postdoctoral Research Fellow at the Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, working with Prof. Laurent Kneip. I also served as a SLAM Technical Consultant at Motovis, leading a team on autonomous driving projects.

I received my Ph.D. from the Australian National University, advised by Prof. Laurent Kneip and Prof. Hongdong Li. I also hold a Bachelor’s degree in Engineering from Beijing Institute of Technology. My research focuses on AI 3D generation, visual SLAM and event-based vision, connecting generative modeling with spatial understanding and robust visual perception.

Computer visionRoboticsSpatial computingGenerative AI

02 / SELECTED WORK

Selected projects.

More on GitHub ↗

03 / RESEARCH

Selected publications.

All publications ↗
Figure 1: StoryBlender overview and comparison with existing workflows.

ECCV 2026, 130–147 · 2026

StoryBlender: Inter-shot consistent and editable 3D storyboard with spatial-temporal dynamics

Bingliang Li, Zhenhong Sun, Jiaming Bian, Yuehao Wu, Yifu Wang, Hongdong Li, Yatao Bian, Huadong Mo, Daoyi Dong

StoryBlender turns a story into editable 3D storyboards while keeping characters, assets and spatial relationships consistent across shots. A persistent scene representation connects narrative planning with engine-verified scene construction.

04 / WHAT'S NEW

Latest news.

Research, releases & milestones.

We have two papers accepted by NeurIPS 2026.

We have one paper accepted by Siggraph-Asia 2026.

We have one paper accepted by ECCV 2026.

We have launched v2fun.ai, a unified AI 3D platform integrating image generation, 3D modeling, and animation into a single workflow.

We have one paper accepted by RA-L 2025.

We have organized workshop on Event-Based Vision for Advanced State Estimation and Image Processing at IROS 2025.

Earlier news ↓

We have released our latest research project: OpenGV 2.0. [Available]

We have released our latest research project: Triplet-Tuning. [Code Available]

We have released our latest research project: Sketch2Scene. [Project Page]

We have one paper accepted by TRO.

We have one paper accepted by IROS 2023.

Our team "XR Penguin" won the 1st place in both Vision Single and Multi Session Categories of Hilti SLAM challenge 2023!

I joined XR Vision Labs of Tencent as a Senior Researcher.

We have one paper accepted by Sensors 2022.

We have two papers accepted by ICRA 2022.

We have one paper accepted by IROS 2021.

I joined SIMIT(Chinese Academy of Sciences) as a Post-Doc researcher.

We have one paper accepted by ICRA 2021.

We have one paper accepted by T-PAMI.

We have one paper accepted by ECCV 2020.

We have one paper accepted by ICRA 2020.

We have one paper accepted by CVPR 2019.

We receive the Best Student Paper Award at ICVS 2017.

LET'S CONNECT

Good ideas start
with a conversation.

Say hello ↗

AROUND THE WORLD

Visitors to this website

View visitor statistics ↗
Flag Counter · Visitor map
Live statistics appear on the published website.