Hiroharu Kato
Tokyo, Japan
Summary
🤩
RockstarHiroharu Kato is a Ph.D. student and ML engineer based in Tokyo with nine years of experience building research-grade computer vision and 3D graphics tools. He is the author and primary code contributor of the Neural 3D Mesh Renderer (CVPR 2018) implementation, where he implemented .obj loaders, geometric primitives like cross-product utilities, "look at" and perspective transforms, and silhouette rasterization. He specializes in bridging research and engineering by integrating models with the Chainer deep-learning framework and adding testing frameworks for reproducibility. Kato combines a research-first mindset with pragmatic engineering, making experimental CV/graphics methods easier to adopt in practice.
10 years of coding experience
Github Skills (14)
mesh-processing10
3d-rendering10
rasterizer10
ras10
rasterize10
rasters10
python10
raster10
chainer10
rasterizing10
numpy9
deep-learning9
computer-vision9
cuda8
Programming languages (2)
Github contributions (5)
hiroharu-kato/neural_renderer
Nov 2017 - Mar 2022
"Neural 3D Mesh Renderer" (CVPR 2018) by H. Kato, Y. Ushiku, and T. Harada.
Role in this project:
ML Engineer
ML EngineerContributions:1 release, 79 commits, 1 PR in 4 years 4 months
Contributions summary:Hiroharu primarily contributed to the core functionality of a neural 3D mesh renderer. They implemented and tested core components for loading and processing 3D object data, specifically focusing on loading `.obj` files, creating cross-product functions, and developing transformations for "look at" and perspective views. Their work also involved the integration and use of the Chainer deep learning framework, indicating a focus on the model's integration with a deep learning pipeline. The user added testing frameworks for the implemented functionality and rasterization of silhouettes for the meshes.
pytorchmeshdeep-learningcomputer-visionkato
hiroharu-kato/style_transfer_3d
Jan 2018 - May 2018
Contributions:12 commits, 8 pushes, 1 branch in 4 months
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.