AIP Acceleration Research

"AIP Acceleration Research" aims to maximize the research result as an AIP Network Laboratory by supporting newly-proposed research projects based on the excellent research results among the laboratory.

AIP Acceleration Research FY 2025

Model-Based Development Platform to Enhance Safety in Autonomous Driving

Grant No.：JPMJCR25U1

Research Director

Takuya Azumi

Professor, Graduate School of Science and Engineering, Mathematics, Electronics, and Informatics, Saitama University

Collaborators

Yutaka Arakawa

Professor, Graduate School and Faculty of Information Science and Electrical Engineering, Kyushu University

Description

This study aims to establish a model-based development (MBD) platform for autonomous driving systems with an emphasis on safety enhancement. By employing MBD, we seek to efficiently design and implement complex software components, thereby reducing development time and costs. Additionally, we plan to advance safety-critical functionalities, such as trajectory generation, and promote the international dissemination of our MBD platform to accelerate the global evolution of autonomous driving technology.

Scalable measurement and utilization of non-cognitive traits

Grant No.：JPMJCR25U2

Research Director

Kensuke Okada

Associate Professor, Graduate School of Education, Department of Integrated Educational Sciences, University of Tokyo

Collaborators

Kyosuke Bunji

Associate Professor, Graduate School of Business Administration, Kobe University

Description

This research aims to develop a social implementation model for a testing technique based on comparative measurements. This approach is designed to efficiently measure non-cognitive traits from a large number of respondents, while controlling for the impact of social desirability bias. To this end, technological developments and empirical research are carried out, in addition to a case study to facilitate its practical application in business and education. Through the reliable measurement of non-cognitive traits, this research seeks to achieve optimal human resource allocation, enhance team building, and allow personalized learning based on individual trait profiles, thereby improving the overall well-being of organizations and society.

Revolutionizing Cardiac Regenerative Medicine: Explorong the Impact of RWD Science in Myocardial Molecular Imaging

Grant No.：JPMJCR25U3

Research Director

Masateru Kawakubo

Assistant Professor, Faculty of Medical Sciences, Department of Health Sciences, Kyushu University

Collaborators

Michinobu Nagao

Associate Professor, Department of Diagnostic Imaging & Nuclear Medicine, Tokyo Women's Medical University

Description

We aim to establish and commercialize a technology that enables simultaneous analysis of myocardial blood flow and motion in a single examination using myocardial SPECT imaging. This technology is envisioned to play a pivotal role in the future, particularly when myocardial cell transplantation derived from iPS cells, a field under active development in Japan, becomes a reality. It is designed to efficiently evaluate improvements in cardiac function in patients after transplantation. By enabling the assessment of post-transplant heart failure improvement in a single examination, this technology has the potential to simplify the implementation of myocardial cell transplantation and accelerate the widespread adoption of cardiac regenerative medicine.

Developing a Virtual Sensing Platform "Vsens" and Healthcare Systems for Orthopedic Disease Estimation and Rehabilitation Support

Grant No.：JPMJCR25U4

Research Director

Yuta Sugiura

Associate Professor, Faculty of Science and Technology, Department of Information and Computer Science, Keio University

Collaborators

Akira Uchiyama	Associate Professor, Graduate School of Information Science and Technology, Department of Information Networking, Osaka University
Koji Fujita	Professor, Center for Medical Innovation, Institute of Science Tokyo
Hiroki Wakatsuchi	Associate Professor, Graduate School of Engineering, Department of Engineering, Nagoya Institute of Technology

Description

We aim to establish "Vsens," a virtual sensing platform designed to support the development of IoT systems involving real-world sensing. Real-world sensing faces challenges such as the burden of preparing training data and trial-and-error sensor placement, and Vsens addresses these issues.
In Vsens, various sensors that exist in the real world are virtually reconstructed, enabling the acquisition of virtual data. The virtual environment replicates different human body shapes, movements, and surroundings, providing output for estimation models applicable to real-world sensing. During the proposed project period, we aim to integrate the Vsens system, develop it into an open-source community, and create innovative healthcare systems.

Development of Techniques for Augmenting Speech-Production Skills through International Challenge Activities

Grant No.：JPMJCR25U5

Research Director

Tomoki Toda

Professor, Information Technology Center, Information Media Division, Nagoya University

Collaborators

Yusuke Yasuda

Project Researcher, National Institute of Informatics, Digital Content and Media Sciences Research Division, Research Organization of Information and Systems

Description

We aim to develop techniques for augmenting speech-production skills capable of converting dynamic speech characteristics depending on how to control speech organs, such as speech and singing expressions, while preserving speech characteristics of the individual's speech organs, such as voice quality, by improving speech processing based on artificial intelligence techniques. We will open a new research area on augmented speech-production skills by holding international challenges related to singing voice conversion that focuses on singing expressions, lead the research area internationally, and yield the progress of basic techniques by activating the research community.

Exploring auxiliary input signals for artificial neural networks: a caveat against data-driven science

Grant No.：JPMJCR25U6

Research Director

Takashi Morita

Designated Senior Assistant Professor, Academy of Emerging Sciences, Chubu University

Collaborators

Takayuki Watanabe

Designated Assistant Professor, Academy of Emerging Sciences, Chubu University

Description

The principal investigator of this project previously found that the performance of recurrent neural networks (RNNs) can be enhanced by feeding them an auxiliary sinusoidal input singnal, which is unrelated to the main data. This discovery raises a cautionary point for data-driven science, which aims to identify the key factors in data through the eyes of AI. This research project seeks to deepen our understanding of how auxiliary input signals influence AI models and to assess the reliability of data-driven approaches to scientific inquiry.

Explainable Automated Program Repair Based On Software Analytics

Grant No.：JPMJCR25U7

Research Director

Norihiro Yoshida

Professor, College of Information Science and Engineering, Ritsumeikan University

Collaborators

Yutaro Kashiwa

Associate Professor, Graduate School of Science and Technology, Science and Technology Major, Nara Institute of Science and Technology

Description

Much research has been done on automating software development tasks using AI, such as automated bug repair using LLM. However, software developers are accountable to stakeholders for the tasks performed by AI. Therefore, this research aims to realize an explainable automated bug-repair technology. Specifically, we will realize an automated bug-repair technique that can output information that provides evidence of the necessity and correctness of the repair.