Scholar
Jacob Zhiyuan Fang
Google Scholar ID: fHWXpq4AAAAJ
Research Scientist, Bytedance/TikTok
Citations & Impact
Citations
1,691
H-index
13
i10-index
16
Publications
20
Co-authors
17
list available
Publications
Resume (English only)
Academic Achievements
- Published paper: Zero-Shot Controllable Image-to-Video Animation (ACM MM'24);
- Led or contributed to multiple video generation projects such as AI Mermaid (Best TikTok AI effect since 2023), Seedance-TikTok version of the foundational video generation model, etc.
Research Experience
- Senior Research Scientist at Global GenAI, ByteDance / TikTok from 2024 to present, developing AI effects from video generative models and collaborating with Bytedance SEED to build state-of-the-art video generative models for TikTok production.
- Applied Scientist at Amazon AGI from 2022 to 2024, working on image/video generation and large-scale diffusion pre-training & post-training.
Background
- A senior research scientist at Bytedance/TikTok, focusing on video generation and multi-modal learning. Recent focus includes controllable video generation, large-scale training for diffusion models, and representation learning.
Miscellany
- Hiring Full-Time employees and long-term interns, looking for highly motivated Ph.D. candidates with experience in video generation/diffusion/AR model training and top-tier publications.
