Summary
Overview
Work History
Education
Skills
Timeline
Generic
Zhongyuan Hu

Zhongyuan Hu

Summary

Experienced video processing algorithm engineer with expertise in diffusion models, and other computer vision technologies. Passionate about AI-generated content, I excel in developing innovative solutions that advance multimodal applications, leveraging AI to transform and enhance video processing.

Overview

5
5
years of professional experience

Work History

Video Processing Engineer

Zoom
11.2021 - Current
  • 3D Avatar: Use 3D avatar to replace the human in meetings, not only generate corresponding avatar from one single image, but also follow the emotions and movements.
  • Text to image: Leverage diffusion models to generate customized image based on user input prompts as well as controls.

Video Processing Engineer

Bigo SG
04.2019 - 10.2021
  • 3D Human Face Reconstruction: apply light-weight CNN to generate 3D faces from one single input image.
  • Virtual VTuber: Present a broadcast by the virtual vtuber avatar by mapping the actions and expressions from a performer.
  • Face Editor: An editor to make face filters including liquify, stretch and so on, with the help of neural networks.

Education

Master of Science - Computer Science

The University of Tokyo
Tokyo, Japan
04.2019

Bachelor of Science - Computer Science

Zhejiang University
Hangzhou, China
06.2016

Skills

  • Diffusion models
  • Generative models
  • Multimodal
  • Transformer
  • Computer vision
  • Computation geometry
  • Differentiable renderer
  • C
  • Python
  • PyTorch
  • OpenGL
  • OpenCV

Timeline

Video Processing Engineer

Zoom
11.2021 - Current

Video Processing Engineer

Bigo SG
04.2019 - 10.2021

Master of Science - Computer Science

The University of Tokyo

Bachelor of Science - Computer Science

Zhejiang University

Similar Profiles

CREATE PROFILE
Zhongyuan Hu