Noguchi Setsuko
Princeton University. Japanese Studies Librarian
The Birth of Speech Synthesis: The Works and Life of Umeda Noriko
Umeda Noriko was one of the pioneering speech scientists who first automated a range of voices – such as male, female, and child – in speech synthesis. She earned a PhD in linguistics from the University of Tokyo and initially worked at the Electrotechnical Laboratory (ETL, 電気試験所). Her significant contributions at ETL led to her recruitment by Bell Telephone Laboratories (BTL). After moving to the US, Umeda and her team developed groundbreaking technologies in the 1970s, including the origins of female voice and natural speech in computer synthesis, as well as audio interfaces for blind individuals. Her remarkable work marks an early era in the history of AI. Later, Umeda became a professor of linguistics at New York University and later director of the Institute for Speech and Language Sciences.
In 2022, Princeton University's East Asian Library received a substantial archival collection of Umeda’s documents, including papers and audiovisual materials recorded in her lab during her tenure at ETL, BTL, and NYU. Despite her outstanding work, from which numerous people have benefited, she has not been widely recognized in Japan. I aim to introduce her by presenting her documented archival records, articles, and books, hoping researchers in the history of science or anthropology will find interest in her work.
音声合成の誕生: 梅田規子の業績と生涯
梅田規子は、音声合成の分野において、初めて男性、女性、子どもなどのさまざまな声を自動化した音声科学者の草分けの一人である。梅田は、東京大学で言語学の博士号を取得し、電気試験所(ETL)で働いた後、ETLでの多大な貢献によりベル電話研究所 (BTL) に採用された。米国に移住した後、梅田と彼女のチームは、コンピュータ合成における女性の音声、自然音声の起源となる画期的な技術や、視覚障害者のためのオーディオインターフェイスなどを1970年代に開発した。AIの歴史の初期における彼女の素晴らしい業績は、特筆すべきものである。その後、梅田はニューヨーク大学の言語学教授、後にSpeech and Language Sciences研究所の所長となり2001年に退職するまで研究を続けた。