Yusuke Uchida
General Manager at Mobility Technologies
Tokyo, Japan
Summary
🤩
Rockstar🎓
Top SchoolYusuke Uchida is a Tokyo-based technology leader and General Manager with nine years of industry experience and a PhD in Information Science and Technology from the University of Tokyo. He blends hands-on machine learning research—especially in deep learning for image recognition, image retrieval, and approximate nearest neighbor search—with product and organizational leadership across Mobility Technologies and DeNA. His open-source work includes notable Keras projects such as a Noise2Noise implementation and real-time age/gender estimation demos, demonstrating both reproducible research and practical deployment skills. Earlier research at KDDI R&D and a visiting fellowship at SRI International reflect a strong background in computer vision and multimedia retrieval. Colleagues value his ability to translate cutting-edge research into production features and demos that drive real user value. He often pairs academic rigor with pragmatic engineering, shipping experiments that evolve into product-ready systems.
9 years of coding experience
12 years of employment as a software developer
Doctor of Philosophy (Ph.D.), Information Science and Technology, Doctor of Philosophy (Ph.D.), Information Science and Technology at 東京大学Master degree of Informatics, Systems science, Master degree of Informatics, Systems science at Kyoto University
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (20)
python10
machine-learning10
mask-rcnn10
keras10
tensorflow10
trainings10
faster-rcnn10
computer-vision10
modeling10
image-processing9
evaluation9
data-generation9
generator9
eval9
dlib9
Programming languages (2)
Github contributions (5)
yu4u/age-gender-estimation
Apr 2017 - Aug 2020
Keras implementation of a CNN network for age and gender estimation
Role in this project:
ML Engineer & Data Scientist
ML Engineer & Data ScientistContributions:1 release, 172 commits, 25 PRs in 3 years 4 months
Contributions summary:Yusuke's contributions centered around the development and utilization of a Convolutional Neural Network (CNN) for age and gender estimation. They implemented and integrated a pretrained model, optimized its parameters, and also created a demo that utilized the developed model for face detection and age/gender prediction in real-time. The user demonstrated skills in data preparation, model evaluation, and integration of a machine learning solution within a practical application.
agedeep-learninggendergender-estimationestimation
yu4u/noise2noise
Jul 2018 - Mar 2019
An unofficial and partial Keras implementation of "Noise2Noise: Learning Image Restoration without Clean Data"
Role in this project:ML Engineer
ML EngineerContributions:1 release, 55 commits, 14 PRs in 8 months
Contributions summary:Yusuke contributed significantly to the Noise2Noise project, implementing and modifying key components for training and testing a denoising model. They focused on the core aspects of the project, including model architecture, loss functions, and data generators, as well as incorporating new features like different noise models and model architectures (U-Net). They also addressed bug fixes and improved the overall training process.
partialrestorationdeep-learningnoise2noiseimage-restoration
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.