Fengyuan Yang

Fengyuan Yang

Final-year Ph.D. candidate in Computer Science

SoC, NUS

Biography

I am currently a final-year Ph.D. candidate at the CVML group of the National University of Singapore (NUS), supervised by Prof. Angela Yao. Before joining NUS, I earned both my Master’s and Bachelor’s degrees at the VIPL research group of the Institute of Computing Technology (ICT), University of Chinese Academy of Sciences (UCAS), under the guidance of Prof. Xilin Chen and Prof. Ruiping Wang.

My research focuses on controllable video generation and human-centric 3D/4D vision. I study how to recover and decouple human motion, camera trajectories, and scenes from videos, and use these factors as independently editable controls to compose coherent videos. Building on this work, I am interested in world modeling and its potential applications to embodied intelligence. Previously, my master’s research explored semantic-guided few-shot learning.

I expect to graduate in early 2027 and am seeking full-time opportunities. Please feel free to contact me!

Download my resumé.

Interests
  • Controllable Video Generation
  • Human-centric 3D/4D Reconstruction
  • World Modeling
  • Embodied Intelligence
Experience

  • Research Intern · Baidu AMU

    Jul 2025 – Sep 2026

Education

  • PhD in CS, 2022-Present

    School of Computing (SoC), National University of Singapore (NUS)

  • MSE in CS, 2019-2022

    Institute of Computing Technology (ICT), University of Chinese Academy of Sciences (UCAS)

  • Exchange Program, 2018

    École Polytechnique Fédérale de Lausanne (EPFL)

  • B.Eng in CS, 2015-2019

    University of Chinese Academy of Sciences (UCAS)

News

[Sep 25, 2026]: ONE-SHOT was accepted to NeurIPS 2026!

[Feb 2026]: HumanBA was accepted to CVPR 2026!

[Jul 2025]: I joined Baidu as a research intern, working on controllable video generation.

[Jun 2025]: Humans-as-Checkerboards was accepted to ICCV 2025!

[Feb 2024]: KITRO was accepted to CVPR 2024!

[Oct 2022]: One paper was accepted to WACV 2023!

[Aug 2022]: I joined CVML research group, SoC, NUS as a PhD. student!

[Oct 2021]: One paper was accepted to WACV 2022!

[Jun 2020]: Winner of “NI” Track in CVPR2020 CLVision Workshop Challenge!

[Sept 2019]: I joined the VIPL research group, ICT, CAS as a Master student!

Publications

ONE-SHOT: Compositional Human-Environment Video Synthesis via Spatial-Decoupled Motion Injection and Hybrid Context Integration
Fengyuan Yang, Luying Huang, Jiazhi Guan, Quanwei Yang, Dongwei Pan, Jianglin Fu, Haocheng Feng, Wei He, Kaisiyuan Wang, Hang Zhou, Angela Yao (2026). ONE-SHOT: Compositional Human-Environment Video Synthesis via Spatial-Decoupled Motion Injection and Hybrid Context Integration. In NeurIPS 2026.

PDF Cite Project Video arXiv

HumanBA: Human-Aware Bundle Adjustment via Global Human-Camera Decoupling
Fengyuan Yang, TANUJ SUR, Tze Ho Elden Tse, Angela Yao (2026). HumanBA: Human-Aware Bundle Adjustment via Global Human-Camera Decoupling. In CVPR 2026.

PDF Cite Code Video

Humans as Checkerboards: Calibrating Camera Motion Scale for World-Coordinate Human Mesh Recovery
Fengyuan Yang, Kerui Gu, Ha Linh Nguyen, Tze Ho Elden Tse, Angela Yao (2025). Humans as Checkerboards: Calibrating Camera Motion Scale for World-Coordinate Human Mesh Recovery. In ICCV 2025.

PDF Cite Code Project Video

KITRO: Refining Human Mesh by 2D Clues and Kinematic-tree Rotation
Fengyuan Yang, Kerui Gu, Angela Yao (2024). KITRO: Refining Human Mesh by 2D Clues and Kinematic-tree Rotation. In CVPR 2024.

PDF Cite Code Poster Video

Semantic Guided Latent Parts Embedding for Few-Shot Learning
Fengyuan Yang, Ruiping Wang, Xilin Chen (2023). Semantic Guided Latent Parts Embedding for Few-Shot Learning. In WACV 2023.

PDF Cite Code

SEGA: Semantic Guided Attention on Visual Prototype for Few-Shot Learning
Fengyuan Yang, Ruiping Wang, Xilin Chen (2022). SEGA: Semantic Guided Attention on Visual Prototype for Few-Shot Learning. In WACV 2022.

PDF Cite Code Poster Slides Video

Awards

ECCV2024 Outstanding Reviewer Award
European Conference on Computer Vision (ECCV) 2024 Sep 2024
CVPR2020 CLVision Workshop Challenge, Winner of “NI” Track
CLVISION CVPR Workshop Jun 2020
Outstanding Undergraduate Award of Beijing
Beijing Municipal Education Commission Jul 2019
Outstanding Undergraduate Award of UCAS
University of Chinese Academy of Sciences Jul 2019
National Student Computer System Capability Challenge, Outstanding winner
Loongson Technology Corp Ltd Sep 2018
The ACM-ICPC Asia Regional Contest EC-Final 2017, Bronze Medal
International Collegiate Programming Contest (ICPC) Dec 2017
The ACM-ICPC Asia Regional Contest Xi’an Site 2017, Bronze Medal
International Collegiate Programming Contest (ICPC) Oct 2017

Earlier Projects

LoCoBot Project
We upgraded our robot to LoCoBot, a mainstream robot for robotics research, and added additional features to this Robot Vision Exploration Project. I designed and maintained the customized communication framework and the demo website.
LoCoBot Project
Loomo Project
In this project we built an intelligent system based on a real robot (i.e., Loomo). I designed and maintained the customized communication framework that transfers data between Loomo and functional modules (e.g., 3D Reconstruction, Scene Graph, etc.) based on ROS.
Loomo Project

Skills

Programming & Frameworks

Python, C/C++
PyTorch, Diffusers

Generative Models

Diffusion / Flow Matching
DiT, Multimodal Conditioning

3D/4D Vision

SMPL/SMPL-X, SLAM
Bundle Adjustment, Camera Geometry

Training

LoRA, DeepSpeed
Multi-GPU Training

Contact