2022 Fiscal Year Final Research Report
Design of Robust Speech Recognition System and Development of its Energy Harvesting System
| Project/Area Number |
18H03212
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 60040:Computer system-related
|
|---|
| Research Institution | Chitose Institute of Science and Technology (2020-2022)
Hokkaido University (2018-2019) |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2018-04-01 – 2022-03-31
|
| Keywords | 音声認識システム / 音声情報処理 / エナジーハーベスト / 低消費電力技術 / 雑音ロバスト / ディジタル信号処理 / ハードウェア・ソフトウェア協調設計 / 回路とシステム |
|---|
| Outline of Final Research Achievements |
In this study, a new and novel speech recognition system has been developed. It is suitable for noisy circumstances and sustainability environment. The developed system has been evaluated under several real noisy circumstances and thus it is realized into an autonomous robot for actual application. In the first step, a new automatic speech recognition system with high performance noise robust has been designed and developed. The expected noise SNR is lower than 10 dB. In order to apply this system into severe noise conditions, noise reduction over the running spectrum domain and psychoacoustic processing have been introduced. At the same time, the system is also considered to be suitable for the sustainability environment. In order words, the low power consumption system design, i.e., the software and hardware co-design method, has been applied to the proposed system and it may be useful for the system with energy harvesting mechanism.
|
| Free Research Field |
情報科学、電子情報通信工学、回路とシステム
|
| Academic Significance and Societal Importance of the Research Achievements |
人工知能の導入により、静かな環境や接話マイクを利用した音声認識は実用化レベルに達している。一方で、ロボットなどとの会話や、屋外などの騒音が気になる環境における音声認識の性能は十分ではなく、雑音環境にロバストな本研究成果が与える社会的意義は大きい。また、エネルギーハーべスティング技術も考慮したシステム設計は、ゼロカーボンを目指すすべての社会において重要な研究テーマであり、学術的意義も大きい。
|
