Yu-guan Hsieh
AI Anime Researcher at Spellbrush
Tokyo, Japan
Summary
🤩
Rockstar🎓
Top SchoolYu-Guan Hsieh is an AI researcher and applied mathematician with nine years of experience specializing in generative models, vision-language pretraining, efficient fine-tuning, and learning in games. After a PhD in optimization and machine learning and top-tier internships (AWS, RIKEN) he held a postdoc at Apple where he led the Graph-Based Captioning project and co-authored multiple 2024 arXiv/ICML papers. Now based in Tokyo as AI Anime Researcher at Spellbrush, he blends rigorous ML research with a creative mission to make anime more realistic using state-of-the-art generative techniques. His work spans theory and practice—from extragradient convergence proofs to building datasets and released code—showing a rare fluency across math, systems, and production research. Notably, he has published on using diffusion priors for bandit decision-making and on practical improvements to flow models and gradient surgery, signaling an aptitude for translating theoretical ideas into applied ML.
9 years of coding experience
1 year of employment as a software developer
Master 2 (M2), MVA Mathematics/Computer Vision/Machine Learning, 18.05/20, Master 2 (M2), MVA Mathematics/Computer Vision/Machine Learning, 18.05/20 at École Normale Supérieure Paris-SaclayEcole normale supérieureDoctor of Philosophy - PhD, Optimization, Machine learning, Doctor of Philosophy - PhD, Optimization, Machine learning at Université Grenoble AlpesBachelor's degree, Master, ENS graduate degree as 'normalien', Computer Science, 17.22/20 and 17.6/20, Bachelor's degree, Master, ENS graduate degree as 'normalien', Computer Science, 17.22/20 and 17.6/20 at École normale supérieureClasse préparatoire, Mathematics, Physics and Computer Science, Classe préparatoire, Mathematics, Physics and Computer Science at Lycée du Parc
Chinese, French, English, Japanese
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (33)
pytorch10
stable-diffusion10
text2image10
generative10
gradio10
deep-learning10
image-generation10
ai10
react10
generative-model9
classification9
flux9
audio9
python8
pipeline8
Programming languages (2)
Github contributions (5)
anthonyhtyao/ocamlTW
Dec 2016 - Oct 2017
Contributions:36 PRs, 64 pushes, 5 branches in 10 months
mandarinocamlpresenting
cyber-meow/anime_screenshot_pipeline
Dec 2022 - Mar 2023
A 99% automatized pipeline to construct training set from anime and more for text-to-image model training
Contributions:61 commits, 82 PRs, 257 pushes in 3 months
pytorchpipelinesemi-automaticdeep-learningsemi
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.