TY - GEN
T1 - Study on method for protecting speech privacy by actively controlling speech transmission index in simulated room
AU - Unoki, Masashi
AU - Kashihara, Yuta
AU - Kobayashi, Maori
AU - Akagi, Masato
N1 - Funding Information:
ACKNOWLEDGEMENTS This work was supported by a Grant-in-Aid for challenging Exploratory Research (No. 16K12458) and Innovative Areas (No. 16H01669) from the Ministry of Education, Culture, Science and Technology (MEXT), Japan, and by the Secom Science and Technology Foundation.
Publisher Copyright:
© 2017 IEEE.
PY - 2018/2/5
Y1 - 2018/2/5
N2 - Protecting speech privacy in a specific room is an important challenge in room acoustics. However, protecting people's conversation from being overheard by an unintended listener, that is, making them not understandable, is difficult. This paper proposes a method for protecting speech privacy by actively controlling the speech transmission index (STI) in a simulated room containing an unintended listener. In this method, the STI in the simulated room can be controlled by manipulating the parameters of the simulated room impulse response (RIR). We can control the STI by convolving speech with the simulated RIR because the presentation of speech and additive delayed-manipulated speech can be regarded as the convolution of speech with late reverberation in the simulated room. Three experiments (world intelligibility, listening difficulty, and annoyance tests) were conducted to compare the proposed method with two conventional methods (noise masking and reverberation). The results showed that speech privacy can be protected by controlling STI derived by manipulating the simulated RIR. The results also showed that the proposed method can protect the privacy of conversations as effectively as those other methods can by using lower noise levels and shorter reverberation.
AB - Protecting speech privacy in a specific room is an important challenge in room acoustics. However, protecting people's conversation from being overheard by an unintended listener, that is, making them not understandable, is difficult. This paper proposes a method for protecting speech privacy by actively controlling the speech transmission index (STI) in a simulated room containing an unintended listener. In this method, the STI in the simulated room can be controlled by manipulating the parameters of the simulated room impulse response (RIR). We can control the STI by convolving speech with the simulated RIR because the presentation of speech and additive delayed-manipulated speech can be regarded as the convolution of speech with late reverberation in the simulated room. Three experiments (world intelligibility, listening difficulty, and annoyance tests) were conducted to compare the proposed method with two conventional methods (noise masking and reverberation). The results showed that speech privacy can be protected by controlling STI derived by manipulating the simulated RIR. The results also showed that the proposed method can protect the privacy of conversations as effectively as those other methods can by using lower noise levels and shorter reverberation.
UR - http://www.scopus.com/inward/record.url?scp=85050807965&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85050807965&partnerID=8YFLogxK
U2 - 10.1109/APSIPA.2017.8282212
DO - 10.1109/APSIPA.2017.8282212
M3 - Conference contribution
AN - SCOPUS:85050807965
T3 - Proceedings - 9th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2017
SP - 1199
EP - 1204
BT - Proceedings - 9th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2017
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 9th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2017
Y2 - 12 December 2017 through 15 December 2017
ER -