G-U-N

Follow

🥹

Out sick

Fu-Yun Wang G-U-N

🥹

Out sick

Follow

Beginner.

285 followers · 33 following

Moon
g-u-n.github.io

Achievements

Achievements

G-U-N/README.md

List of Projects

Diffusion Acceleration

The Principle and Application of Diffusion Acceleration: My talk on diffusion acceleration related research.
Stable Consisrency Tuning: Understanding and improving consistency models.
Rectified Diffusion: Understanding and extending rectified flows to general diffusion models.
PCM [NeurIPS 2024]: Phasing and improving latent consistency models (One-step T2I and T2V).
OSV: Efficient Image-to-Video Generation.

Video Editing/Synthesis

Motion-I2V [SIGGRAPH 2024]: General explicit motion generation framework, working for Image-to-Video, Drag Video, Motion Brush, Vid2Vid.
AnimateLCM [SIGGRAPH ASIA 2024 TCS]: Efficient personalized video generation without personalized video data.
Be-Your-Outpainter [ECCV 2024]: One-shot video outpainting.

Class-Incremental Learning

PyCIL: a python toolbox for class-incremental learning.
CIL.Survey: a survey in deep class-incremental learning.
CIL.Pytorch: a pytorch tutorial to class-incremental learning.

Contact

email: wangfuyun(at)smail.nju.edu.cn; fywang(at)link.cuhk.edu.hk; fywang0126(at)gmail.com

Pinned Loading

PyCIL PyCIL Public

PyCIL: A Python Toolbox for Class-Incremental Learning

Python 855 139
Rectified-Diffusion Rectified-Diffusion Public

Rectified Diffusion: Straightness Is Not Your Need

Python 153 5
Motion-I2V Motion-I2V Public

[SIGGRAPH 2024] Motion I2V: Consistent and Controllable Image-to-Video Generation with Explicit Motion Modeling

Python 127 11
Phased-Consistency-Model Phased-Consistency-Model Public

[NeurIPS 2024] Boosting the performance of consistency models with PCM!

Python 399 15