Masaaki Iiyama

Data Oriented Real-time Information Analysis Platform for Sustainable Fishery

Research Director

Masaaki Iiyama

Collaborator

Yuusuke Tanaka

Visiting Engineer Center for Earth Information Science and Technology Japan Agency for Marine-Earth Science and Technology

Outline

Our goal is to develop a data oriented real-time information analysis platform for sustainable fishery, which is both environmentally and economically efficient. In our project, we accumulate various big data that relates to ocean weather and fishery, and develop AI technologies that analyze the big data. The technologies include high-precision sea weather forecasting and estimation of good fishing spot from ocean big data.

Takenao Ohkawa

Innovation in management of breeding cows in pasture by interaction analysis

Research Director

Takenao Ohkawa

Collaborator

Kenji Oyama

Professor Graduate School of Agricultural Science Kobe University

Outline

This project aims at bringing an epoch-making innovation enabling labor saving and cost reduction in management of breeding cows in pasture to Japanese beef farmers. We will develop novel technologies for extraction, analysis and interpretation of various types of information about interactions between cows and establish a framework for the integration of them to detect weak expression of the state of estrus and physical/mental health by focusing on the social nature of cows that can be observed through interaction analysis.

Shimpei Kato

Risk and Anomaly Prediction in Fully Autonomous Driving

Research Director

Shimpei Kato

Collaborator

Kenzi Kono	Professor Faculty of Science and Technology Keio University
Kazuya Takeda	Professor Institute of Innovation for Future Society Nagoya University

Outline

This research contributes to production of autonomous driving systems that continuously improve intelligence by run after run. The scope of intelligence includes not only a basic automation capability, such as perception, planning, and control, but also a prediction capability for risk of driving scenes and anomaly of the running system. We believe that risk and anomaly prediction is becoming the most significant capability to ensure safety and comfort of emerging autonomous driving technology. This research develops its platform.

Taishiro Kishimoto

Understanding Psychiatric Disorders Through Natural Language Processing: A New Approach for Prevention and Early Diagnosis

Research Director

Taishiro Kishimoto

Collaborator

Yoshinobu Kano

Associate Professor College of Informatics Shizuoka University

Outline

Psychiatric disorders such as depression, schizophrenia and dementia are diagnosed through patients’ “language”. In order to choose accurate treatments, it is crucial to extract the characteristics of these diseases and understand them thoroughly through patients’ language. In this project, we will be utilizing a natural language processing-based approach to digitize/quantify disease characteristics. By doing so, we will expand our understanding of the diseases and eventually lead to better prevention and/or early diagnosis of psychiatric disorders.

Katsufumi Sato

Cyber ocean: next generation navigation system on the sea

Research Director

Katsufumi Sato

Collaborator

Yukio Komatsu	Associate Professor Graduate School of Frontier Sciences The University of Tokyo
Kaoru Saito	Professor Graduate School of Frontier Sciences The University of Tokyo

Outline

To understand physical interaction between the atmosphere and ocean, and for more accurate weather forecast, in situ measurements are essential. Now a day satellite remote sensing systems can provide information on water temperature, current, wind and wave. Combining these data in a supercomputer, numerical models nowcast/forecast atmosphere and ocean environments. Improvement in the accuracy of the numerical models requires additional measurements with higher spatial and temporal resolutions. However, measurements by existing satellite remote sensing and meteorological buoys are coarse spatially and temporally. Simultaneous measurements using several research vessels and buoys cost a lot to cover wider range of ocean. Here we propose a novel system using marine top predators such as seabirds, turtles and cetaceans as living meteorological buoys to monitor pelagic environments. We will deploy small recorders on them to record their 3 dimensional tracks in air and water, with physical environments. Fine-scale data collected by the animals can provide us valuable information to fill the gaps in terms of both time and space. As a result, we can expect to contribute for improvement in resolution and accuracy of the numerical simulation for nowcast/forecast environment boundary between atmosphere and ocean.

Satoh Shin’ichi

Experience and Action Sensing of Media Consumers based on Unknown Target Retrieval and Recognition Framework

Research Director

Satoh Shin’ichi

Collaborator

Kiyoharu Aizawa	Professor Graduate School of Information Science and Technology The University of Tokyo
Toshihiko Yamasaki	Associate Professor Graduate School of Information Science and Technology The University of Tokyo

Outline

We first develop unknown target retrieval and recognition technologies to detect significant changes and trends in dynamic media including broadcast videos, social network services, and lifelogs. Based on the technology suite, we will build a framework to sense how people obtain information from broadcast videos and SNS and how people react following the obtained information. The framework will enable early detection of new trends such as brand new products, analysis of effective marketing strategies raising buying behavior, analysis of mechanism driving people for humanitarian behavior, and so on.

Koichi Shinoda

Fast and cost-effective deep learning algorithm platform for video processing in social infrastructure

Research Director

Koichi Shinoda

Collaborator

Satoshi Matsuoka	Specially Appointed Professor School of Computing Tokyo Institute of Technology
Tsuyoshi Murata	Associate Professor School of Computing Tokyo Institute of Technology
Rio Yokota	Associate Professor Global Scientific Information and Computing Center Tokyo Institute of Technology

Outline

We aim to establish a high-performance real-time deep learning algorithm basis for detecting objects and anomalies from a large amount of high definition videos recorded by drive recorders, surveillance cameras, and the like. Computer science researchers specializing in four different levels from architectures to applications including GPU fast computation, parallel computation, machine learning, and data mining, collaborate to realize 1000 times faster processing with 0.1% the amount of memory compared to conventional platforms.

Goichiro Hanaoka

Universal Server-aided Computation for Realizing Secure Private Data Processing

Research Director

Goichiro Hanaoka

Collaborator

Kiyoshi Asai

Professor Graduate School of Frontier Sciences The University of Tokyo

Outline

In this research, we aim at realizing various information services while preventing sensitive information from being exposed. Specifically, we unify state-of-the-art cryptographic tools to develop a universal server-aided computation system which enables various types of data processing to be done while keeping the data confidential. Based on this technology, we aim at enabling a “user-friendly society” in which privacy-preserving services can be provided to individuals e.g. AI-based medical checkup.

Ryuji Hamamoto

Project on Development of Integrated Medical System for Diagnosis and Treatment of Cancer by Artificial Intelligence

Research Director

Ryuji Hamamoto

Collaborator

Daisuke Okanohara	Vice President Preferred Networks Inc.
Jun Sese	President ＆ CEO Headquarters Humanome Lab

Outline

In this study, we plan to analyze the multi-omics data stored in National Cancer Center such as genomic data, epigenetic data, medical imaging data and microRNA data, clinical information and literature information in an integrated manner by artificial intelligence. Subsequently, we will develop the integrated medical system based on our achievements for Japanese cancer patients. Furthermore, we will widely distribute the medical system in a society in order to improve the quality of medical care for cancer patients.

Shiho Moriai

Privacy-preserving Data Analytics to Promote Cross-industry Data Sharing

Research Director

Shiho Moriai

Collaborator

Seiichi Ozawa	Professor Center for Mathematical and Data Sciences Kobe University
Takahiro Sugawara	CEO Eltes Co.,Ltd.

Outline

In sharing and analyzing big data across multiple industries and organizations, there are obstacles and challenges on how to protect data security and privacy. Employing artificial intelligence in combination with cryptographic technologies, this research project induces technological innovations for big data analytics with the development of efficient machine learning algorithms for encrypted data. We aim at applying our research results to real problems in financial industry such as detecting illegal money transfers and making credit limit decisions based on big data across industries.