Takeru Miyato
Doctoral Student at ELLIS - European Laboratory for Learning and Intelligent Systems
Netherlands
Summary
🤩
Rockstar🎓
Top SchoolTakeru Miyato is a Ph.D. student at the Autonomous Vision Group, University of Tübingen and a part-time researcher at Preferred Networks with a decade of experience bridging academic research and industrial ML engineering. His work focuses on generative models and spectral normalization in GANs—evidenced by substantial contributions refactoring generators and discriminators in the well-known pfnet-research sngan_projection repository. He has a history of high-impact internships and research roles, including a Google Brain internship under Ian Goodfellow and contributions at Preferred Networks since 2016. Comfortable in both deep research and hands-on code hygiene, he blends rigorous theoretical training from Kyoto University with practical engineering that ships robust model architectures. Based in the Netherlands, he routinely collaborates internationally, having held visiting positions at Caltech and an exchange at the University of Amsterdam.
10 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD, Computer science, Doctor of Philosophy - PhD, Computer science at Eberhard Karls Universität TübingenBEng in EE & MSc in Informatics, BEng in EE & MSc in Informatics at Kyoto University
Github Skills (10)
machine-learning10
deep-learning10
cgan10
cyclegan10
dcgan10
image-generation9
refactoring8
faster-rcnn8
mask-rcnn8
refactor8
Programming languages (2)
Github contributions (5)
pfnet-research/sngan_projection
Feb 2018 - Oct 2018
GANs with spectral normalization and projection discriminator
Role in this project:
ML Engineer
ML EngineerContributions:41 commits, 4 PRs, 50 pushes in 8 months
Contributions summary:Takeru primarily contributed to the project by modifying and refactoring the generator and discriminator models. Their work involved renaming and restructuring generator classes, fixing spectral normalization links, and adjusting default settings to align with the project's objectives, particularly those related to image generation with spectral normalization. Additional contributions include bug fixes, code style improvements, and adding new model architectures (e.g. 256 models) to the project.
pytorchprojectiondeep-learningspectral-normalizationgenerative-adversarial-network
jphacks/KB_1502
Nov 2015 - Dec 2015
Contributions:62 commits, 38 pushes, 6 branches in 14 days
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.