PKSHA Technology Initiates Joint Research with Nara Institute of Science and Technology (NAIST) on Next-Generation Speech Synthesis Technology

PKSHA Technology, Inc. (Read: Parksha Technology, Headquarters: Bunkyo-ku, Tokyo, Representative Director: Katsuya Uenoyama, hereinafter PKSHA) is pleased to announce that it will commence joint research on next-generation speech synthesis technology with the National University Corporation, Nara Institute of Science and Technology (Location: Ikoma-shi, Nara, President: Kazuhiro Shiozaki, hereinafter NAIST) starting June 2026.

This joint research aims to establish next-generation speech synthesis technology capable of generating natural and rich speech expressions even from limited voice data. Specifically, we will focus on research and development of speech generation methods for flexibly controlling emotions and nuances of speech without compromising speaker individuality. Through this, we aim to realize diverse and rich voice communication in applications such as conversational AI, avatars, and voice agents, and to accelerate the social implementation of next-generation voice AI.

Background and Objectives of the Joint Research

PKSHA has provided algorithmic solutions utilizing natural language processing, speech recognition/synthesis, machine learning, and deep learning technologies, promoting the social implementation of AI with many companies. On the other hand, NAIST has been promoting cutting-edge research in the field of speech information processing, particularly in information science, contributing to the development of fundamental technologies related to speech synthesis and recognition.

This joint research, themed "Establishment of next-generation speech synthesis technology enabling natural and rich speech expressions from small amounts of voice data," aims to accelerate the creation and practical application of new speech synthesis technologies by fusing PKSHA's R&D capabilities rooted in social implementation with NAIST's advanced research and development capabilities in speech information processing. Through this joint research, we will advance the research and development of speech generation technology that flexibly controls emotions and speaking styles without compromising speaker individuality, aiming to realize diverse and rich voice communication and accelerate the social implementation of next-generation voice AI.

PKSHA's Achievements in Speech and Natural Language Processing

PKSHA has consistently promoted research and development of fundamental technologies and their deployment into practical services in the fields of natural language processing and speech processing. For example, through research and development of accent estimation technology that contributes to natural speech generation in Japanese speech synthesis, and by open-sourcing the results, we have contributed to the advancement of fundamental technologies in the speech synthesis field. We are also promoting the provision of solutions utilizing speech recognition, speech synthesis, and speech analysis technologies, working to promote the utilization of voice data in practical environments, including the contact center domain. The know-how cultivated through these initiatives in voice data expression design, model optimization, and evaluation/improvement in operational environments will form the basis for technology development with a view to social implementation in this joint research.

Nara Institute of Science and Technology's Achievements in Speech Processing

At NAIST, cutting-edge research in speech information processing is being promoted, centered on the field of information science. In particular, the Human-AI Interaction Lab (HAI Lab) engages in a wide range of themes from basic to applied research related to speech generation, recognition, and dialogue, accumulating achievements in areas such as speech expression modeling, analysis of the relationship between speaker characteristics and speech expression, and research on dialogue systems. Furthermore, NAIST has received international recognition through research presentations and paper acceptances at academic conferences both domestically and internationally, contributing to the development of fundamental technologies in the field of speech information processing.

About Nara Institute of Science and Technology

Nara Institute of Science and Technology is a national graduate university with a focus on three fields: Information Science, Bioscience, and Materials Science, promoting advanced research and human resource development in cutting-edge science and technology. In the field of Information Science, in particular, it engages in advanced research in areas such as artificial intelligence, speech information processing, and natural language processing, and is highly regarded in domestic and international research communities.

University Name: National University Corporation Nara Institute of Science and Technology

Location: 8916-5 Takayama-cho, Ikoma-shi, Nara

Representative: President Kazuhiro Shiozaki

URL: https://www.naist.jp/

Company Profile of PKSHA Technology, Inc.

Under the mission "Shaping the Future of Software," we provide diverse AI solutions that solve societal challenges. By deploying these as "AI Solutions" optimized for various industries such as finance, manufacturing, and education, and as versatile "AI SaaS" such as "PKSHA AI Helpdesk," "PKSHA ChatAgent," and "Interview Copilot," we support the future of work and realize a society where humans and software evolve together.

Company Name: PKSHA Technology, Inc. Location: Hongo Segawa Building 4F, 2-35-10 Hongo, Bunkyo-ku, Tokyo Representative: Katsuya Uenoyama, Representative Director URL: https://www.pkshatech.com/

Contact for Inquiries: [email protected] Keywords: