Work Experience

Narrative Nights Inc. (2011/Jul. ~)

CEO / Lead programmer


AR / MR / 3D

Massive 3D security data visualizer

Massive 3D security data visualizer client implementation.
(Nirvana Kai, Daedalus, NICTER Atlas/Cube)

Client NICT, The Strippers
Platform WebGL/iOS/Android
Environment Unity, C#, Shader


AR Music Kit for Google I/O 2016 by Yuri Suzuki

AR development

https://www.androidexperiments.com/experiment/ar-music-kit

Client Yuri Suzuki, Gold Rush Computing
Platform Android
Environment Java, C++, ARToolkit


Hololens app

Hololens app implementation for the museum and historical sites with the light control.
(Details undisclosed.)

Client Undisclosed (Ad agency)
Platform Hololens, Stage lighting
Environment Unity, C#, DMX512


AKB48 AR

Real photo AR develpment with hemishpere shooting system.

Client MTI Ltd.
Platform Android, iOS
Environment C++, Objective-C, Java, Vuforia


AI / ML

Asychronous learning algorithm implementation

Asynchronous learning algorithm impementation for Post-K (FUGAKU) project.

Client RIKEN
Environment Python, TensorFlow, BriCA


Reinforcement Learning implementation

Reinforcement learning demo with CARLA/Uneal Engine

Client Undisclosed
Platform Linux
Environment UnrealEngine, CARLA, C++, Python, TensorFlow


fMRI/EEG data analysis

fMRI/EEG simultaneous recording data analysis with deep learning.

Client Kochi University of Technology
Environment Python, PyTorch


Image Processing / DSP

Picute

GPU filter composition system development

Client MTI Ltd.
Platform Android, iOS
Environment Java, C++, Objective-C, GLSL


Pocket Video

GPU based video editing system development.
DSP implementation.

Client Pocket Supernova
Platform Android, iOS
Environment Java, C++, Objective-C, GLSL


Interactive

JOC Official App

London Olympic JOC official social app: Android client development

Client JOC, The Strippers
Platform Android
Environment Java


Miidas HR Science Institute (2020/10. ~ )

AI Programmer

Recommendation system development

Environment PyTorch, Catboost, gRPC


Dwango AI Lab (2018/6. ~ 2019/3.)

Cooperative researcher

Reseach of disentanglement representation

Partner NTT Laboratory
Environment Python, PyTorch, TensorFlow

Reinforcement learning environment development

Environment Python, OpenGL, OpenCV, React


Cyber Step Inc. (2005/Aug. ~ 2011/Jun.)

Lead programmer

Cosmic Break

Server side development
Game client/graphics development
P2P Voice chat system devepment

Platform Windows, Linux
Environment C++, Java, OpenGL


GetAmped2

Sound streaming server/client development

Platform Windows, Linux
Environment C++, Java


Monemone

Server/Client development

Platform Web
Environment Java, PHP, Squirrel


Sony Computer Entertainment Inc. (2000/Jul. ~ 2004/Dec.)

Game Programmer


ALIFE prototype

Developed genetic programming learning & rigid body simulation system.

Platform PlayStation2
Environment C++, Assembler


DJ Box

Auto DJ playing system (AI-DJ) and sound visualizer development

Platform PlayStation2, PSX
Environment C++


Toro to Kyujitsu

Chatterbot script programming

Platform PlayStation2
Environment Original DSL


Rimococoron

Production tool development

Platform PlayStation2
Environment C++, Java



Open source projects

UNREAL

Reproducing DeepMind paper “Reinforcement Learning with Unsupervised Auxiliary Tasks” with TensorFlow & DeepMindLab.

https://github.com/miyosuda/unreal

Environment Python, TensorFlow


A3C

Reproducing DeepMind paper “Asynchronous Methods for Deep Reinforcement Learning” with TensorFlow.

https://github.com/miyosuda/async_deep_reinforce

Environment Python, TensorFlow


Rodentia

Customizable 3D learning environment for MacOSX and Linux.

https://github.com/miyosuda/rodentia

Platform Linux, MacOSX
Environment C++, OpenGL, GLSL, Python


SCAN

Reproducing DeepMind paper “SCAN: Learning Abstract Hierarchical Compositional Visual Concepts.”

https://github.com/miyosuda/scan

Environment Python, TensorFlow


Disentangled VAE

Reproducing DeepMind paper “Early Visual Concept Learning with Unsupervised Deep Learning.”

https://github.com/miyosuda/disentangled_vae

Environment Python, TensorFlow


Heart rate variability analysis tool

https://github.com/miyosuda/heartrate-monitor

Heart rate variability (HRV) analysis tool to detect autonomic nerve state using auto regressive method.

Platform MacOSX, Polar H7/H6
Environment Swift, Bluetooth LE



Awards

Animal AI Olympics 2019

Category Prize for internal representation category (inf mnky team)

Environment Unity, Python, TensorFlow, Unity-ML Agents



Education

Bachelor of Liberal Arts (Natural Science, Simulation Physics), International Christian University, Tokyo (1999)

Thesis: “1/f fluctuation on the edge of chaos”



Publications

Masanori Yamada, Kim Heecheol, Kosuke Miyoshi, Tomoharu Iwata, Hiroshi Yamakawa (2020)
Disentangled Representations for Sequence Data using Information Bottleneck Principle
ACML, 2020

Masanori Yamada, Kim Heecheol, Kosuke Miyoshi, Hiroshi Yamakawa (2019)
FAVAE: Sequence Disentanglement using Information Bottleneck Principle
arXiv, 2019

Kosuke Miyoshi, Naoya Arakawa, Hiroshi Yamakawa (2018)
Do top-down predictions of time series lead to sparse disentanglement? P15-16 [Poster]
The 28th Annual Conference of the Japanese Neural Network Society, 2018

Kosuke Miyoshi, Naoya Arakawa, Hiroshi Yamakawa (2018)
Hierarchical Variational Recurrent Autoencoder with Top-Down prediction.
Poster session: Joint symposium of 10th Optogenetics Research Conference and Second International Symposium on Brain Information Dynamics 2018

Kim Heecheol, Masanori Yamada, Kosuke Miyoshi, Hiroshi Yamakawa (2018)
Disentangled Representation Learning From Sequential Data.
The 32nd Annual Conference of the Japanese Society for Artificial Intelligence, 2018



Certificates

Coursera Machine Learning, Stanford University Online

Coursera Bayesian Statistics: From Concept to Data Analysis, University of California, Santa Cruz Online