From IT infrastructure to application technology: make, connect, use

The rapid growth of devices (smartphone, tablets, robots, wearables, etc.) and the Internet has increased the amount of information that is being produced and accessed by society. In order to better utilize the data produced from millions of devices and systems, we are conducting research and development in a wide range of fields at the interface between information technologies and human factors. Our mission is to engage and enrich the public through the research and development of intelligent systems combining computational and physical capabilities for human use. A key component of our mission is making new discoveries in the hardware and software that interacts with physical devices to sense and change the state of the real world. Our discoveries will lead to industry innovations and contribute to the advancement of society by facilitating the interaction of humans with cyber-physical systems.

New Research Results

Development of a SAR Foundation Model Specialized for Japan's Land Area Based on Observation Data Obtained from ALOS-2 Satellite

Researchers at AIST, in collaboration with JAXA, have developed a national land-specific Synthetic Aperture Radar (SAR) foundation model using high-resolution observation SAR data routinely acquired over Japan by the SAR instrument, PALSAR-2, onboard the "DAICHI-2" (ALOS-2).
The PALSAR-2 observation data covers the entire country, and by extracting image patches centered on selected locations, a training dataset reflecting diverse land use and land cover types was constructed. Using this dataset, we performed large-scale self-supervised learning to build a SAR foundation model specialized for Japan’s land area. The results from transfer learning for land use/land cover estimation showed significantly higher accuracy compared to models trained from scratch, demonstrating the effectiveness of the foundation model trained on a more balanced dataset. Although the foundation model is initially developed by researchers at AIST and JAXA, it is expected to broaden the use of SAR by utilizing the foundation model that can reduce both the cost of model development and the barrier to interpreting SAR images, which typically requires specialized expertise.
This research was conducted based on the "Agreement on Research and Development of AI Analysis Methods for Satellite Data" between AIST and JAXA, and AIST policy-based budget project "R&D on Generative AI Foundation Models for the Physical Domain". We used ABCI 3.0 provided by AIST and AIST Solutions with support from “ABCI 3.0 Development Acceleration Use”. The details of the research results were presented at the 78th Annual Conference of the Remote Sensing Society of Japan (Spring 2025) held from June 4 to June 5, 2025.

Construction and anticipated applications of a national land-specific SAR foundation model. The high-resolution land use and land cover map of Japan, used in this work, was provided by JAXA.

The Release of Japanese Speech Foundation Models "Izanami" and "Kushinada"

Researchers at AIST has released two Japanese speech foundation models, "Izanami" and "Kushinada," which can be used to build high-performance speech AI.
A speech infrastructure model is a general-purpose AI model for processing and analyzing speech data, which is increasingly being applied to speech recognition and speech emotion recognition. Building a speech foundation model requires at least a few thousand hours of speech data, which is based on the target language and the scene in which it is used. However, speech data such as conversational speech is scarce compared to single-talker speech, and speech AI performance has been insufficient for conversational speech that includes emotionally rich expressions.
We have built and released two Japanese speech foundation models, "Kushinada" and "Izanami," using the largest scale of Japanese speech data ever for creating a foundation model, 60,000 hours. The models are named after Japanese mythology in the hope that they will serve as creators and supporters of various types of speech AI in the future.
“Izanami" can be easily fine-tuned using user data, and "Kushinada" shows high performance in Japanese speech emotion recognition and speech recognition. These features enable the construction of high-performance speech AI even when only a small amount of labeled data is available, such as in the case of elderly people's speech or conversations containing emotionally rich expressions. In the future, we will work on improving speech recognition performance for Japanese dialects. It is expected to be used in a variety of situations, such as improving the problem of poor performance of speech AI due to regional and generational differences, and taking minutes in local assemblies.
The model can be downloaded from the AI model publishing platform Hugging Face (https://huggingface.co/imprt).

Construction of speech AI using the Japanese speech foundation models "Izanami" and "Kushinada
Improved the performance of speech AI, which had been limited by the small amount of training data, by using the feature representation of speech obtained from the Japanese speech foundation model.

Open Innovation Laboratory

Since FY 2016, as a part of the “Open Innovation Arena concept” promoted by the Ministry of Economy, Trade and Industry (METI), AIST has created the concept of “open innovation laboratories” (OILs), collaborative research bases located on university campuses, and has been engaged in their provision. We are planning to establish more than ten OILs by FY 2020.

AIST will merge the basic research carried out at universities, etc. with AISTʼs goal-oriented basic research and applied technology development, and will promote bridging research and evelopment and industry by the establishment of OILs.

Cooperative Research Laboratories

In order to conduct research and development more closely related to strategies of companies, we have established collaborative research laboratories, bearing partner company names.

Partner companies provide their researchers and funding, and AIST provides research resources, such as its researchers, research facilities, and intellectual property. The loaned researchers of companies and AIST researchers jointly conduct research and development.

By setting up cooperative research laboratories, we will accelerate the commercialization of our goal-oriented basic research and application research with partner companies.