AI Model Verification for Intellectual Property Protection
We study model ownership verification, watermarking, authenticity checking, and detection of model theft.
Learn More →We conduct research on security foundations for a safe, trustworthy, and sustainable society in which humans and AI coexist.
Security Infrastructure for AI Society Laboratories address emerging security challenges caused by the rapid deployment of AI systems, including generative AI, federated learning, edge AI, AI hardware, and post-quantum cryptography.
As AI becomes part of social infrastructure, society must address model theft, unauthorized learning, adversarial inputs, privacy leakage, side-channel attacks, and implementation vulnerabilities. Our laboratories study both “security for AI” and “AI for security” from software, hardware, and cryptographic perspectives.
Our laboratories cover a broad range of security infrastructure technologies for AI society, from AI model protection to cryptographic hardware implementation.
We study model ownership verification, watermarking, authenticity checking, and detection of model theft.
Learn More →We develop platforms for privacy-preserving, fair, attack-resistant, and practical federated learning.
Learn More →We investigate efficient methods for removing specific users or data from trained federated models.
We study technologies to prevent and detect unauthorized use of data and content in generative AI training.
We analyze adversarial examples, natural noise, physical-world attacks, backdoors, and data poisoning.
Learn More →We study PUF-based device authentication, hardware fingerprints, secure key generation, and IoT security.
Learn More →We use deep learning to evaluate information leakage from cryptographic implementations and improve security.
Learn More →We study energy-efficient AI processors, FPGA-based acceleration, edge AI, and secure AI hardware.
Learn More →We study CRYSTALS-Kyber, IoT-oriented implementation, energy-harvesting devices, and side-channel resistance.
We study security modules based on lightweight cryptography for resource-constrained IoT devices and embedded systems.
Learn More →We develop secure IoT modules that operate without batteries by using energy harvesting technologies for sustainable deployment.
Learn More →Research outcomes from our laboratories are published in journals, international conferences, and domestic conferences. We have also successfully obtained numerous competitive research grants, including Grants-in-Aid for Scientific Research This section focuses on international conference publications.
Our laboratories consist of multiple faculty members and students working on AI security, privacy-preserving learning, cryptography, and hardware security.
Faculty of Information Engineering, Meijo University
Professor: Masaya YOSHIKAWA
Associate Professor: Yusuke NOZAKI
Assistant Professor: Shu TAKEMOTO
Doctoral and master's students conduct research toward international publication and real-world deployment. Students regularly present their work at domestic and international conferences and publish papers in peer-reviewed academic journals.
Undergraduate students participate in research projects and develop fundamental skills for advanced study.
We welcome motivated applicants for the following graduate programs who wish to contribute to the development of secure and trustworthy AI technologies for society.
We are currently recruiting students for the Master’s Program and Doctoral Program.
Prospective graduate students who are interested in AI security, trustworthy AI, privacy-preserving machine learning, hardware security, and cryptographic implementation are encouraged to contact us through the official inquiry form. Our laboratories encourage students to publish papers in international conferences and journals.
We welcome international students who are interested in conducting research in AI security and related fields.
Master's students are required to complete graduate-level coursework in addition to their research activities. Courses offered in the graduate program can be taken in English, enabling international students to pursue their studies and research effectively.
We actively welcome applicants from other universities, departments, and academic backgrounds.
These examples are representative backgrounds, but applicants are not limited to these fields. Students from diverse academic disciplines are encouraged to apply.
We welcome working professionals who wish to pursue a master's or doctoral degree while continuing their careers.
Research plans and meetings can be arranged flexibly according to each student’s working schedule.
For laboratory visits, admission inquiries, research collaboration, and media inquiries, please use the inquiry form.
To reduce automated email address collection and spam, direct email addresses are not displayed on this website.