Privacy protection for speech signals
Kazumasa Yamamoto, Masatoshi Tsuchiya, Seiichi Nakagawa · Procedia - Social and Behavioral Sciences · 2010
In this paper, we describe about some privacy protection techniques for speech signals captured by public sensors. Speech signal includes the privacy information, such as “voice characteristics” and “linguistic privacy information.” We try to protect the privacy information by using “voice conversion” and “deletion of privacy linguistic information from speech recognition result.” Additionally, “elimination of speech in noisy speech” technique is also considered.