Development of Open Corpus and AI-based Speech Restoration System for Archiving Historical Audio
- Project Leader
- TAKAMICHI Shinnosuke (The University of Tokyo)
- Project Period
- October 2022 -
Summary
The aim of this research is to develop an open speech corpus and automatic speech restoration system for archiving historical audio.
Several historical spoken languages have deteriorated over time, and they contain many acoustic distortions.
In this project, a speech corpus for the analysis of the Japanese language and a deep-learning technique for speech restoration are developed using the dialect of the Tohoku region.
After the end of the project period, the corpus and restoration techniques will be made publicly available free of charge for research purposes only.
