Hey! This is Haozhou’s (昊洲) page. I now work at Bytedance as a machine learning engineer in Shenzhen, China. My areas of interest span over Computer Vision, LLM, Computer Graphics, Theoretical CS, and competitive programming.
Before joining Bytedance, I did my undergrad and M.Sc in Alberta, Canada, both in Computing Science. During my Master program, I was fortunate enough to work with Dr. Mohammad Salavatipour.
Best way to contact me: haozhoupang PUT THE SIGN HERE gmail DOT com.
Fun Projects
- Motion Capture from Video - “JiNiTaiMei”. Extracting joints rotation and root translation from videos then retargeting to a customized character. Algorithm runs in real-time.
- Audio-driven Dance Generation. We develep an global position aware large model to generate realistic and diverse group dances while maintaining strong music correlation and dancer-wise consistency.
- Emotional Talking Face Generation. We build a model to generate realistic 3D talking face conditioned on audio. The model supports multiple languages and 8 different emotions. (happy, sad, etc.) The model runs in real-time with RTF 0.1.
- Garments Simulation. We build a model that takes character’s skeleton infomation to predict the cloth and hair animation. We make the model small enough to be deployed on low-rank devices and achive single frame inference time less than 0.3ms. Players that use mobile devices with limited computing power can also experience realistic garments animation that usually only available for PC.
Education
M.Sc. in Computing Science, University of Alberta, GPA: 3.9/4.0
2018.09 - 2020.11, Edmonton, Canada
- Outstanding Thesis Award
B.Sc in Computing Science, University of Alberta, Graduated with First Class Honors
2014.09 - 2018.6, Edmonton, Canada
- the winner of Dean’s Silver Medal
- University of Alberta Academic Excellence Scholarship
- Dean’s Honor Roll Student
- A Faculty of Science Academic Excellence Scholarship
- A University of Alberta China First Year Excellence Scholarship
- An International Student Scholarship