| Presentation | 2015-11-12 Intelligibility prediction method for outdoor public address systems using classifiers Yosuke Kobayashi, Kazuhiro Kondo, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this paper, we have proved a subjective evaluation and prediction of speech intelligibility using the outdoor public address system of 10 locations for which impulse responses were recorded in Sendai city after the Great East Japan Earthquake. First, the results of the subjective intelligibility evaluation by different test word lists with the same sound field conditions showed the root mean squared error (RMSE) to be very small, as little as 9.2% at maximum. Next, we generated the intelligibility prediction model trained with true/false results of 13 subjects using the support vector machine (SVM). We evaluated the predictive performance of the prediction model using the data that was not used in training, and the RMSE was 4.1%. This result was shown to be highly accurate, even less than the subject experiment result. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | outdoor public address system / speech intelligibility / intelligibility prediction / machine learning / SVM |
| Paper # | EA2015-26,EMM2015-47 |
| Date of Issue | 2015-11-05 (EA, EMM) |
| Conference Information | |
| Committee | EA / EMM |
|---|---|
| Conference Date | 2015/11/12(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Kumamoto Univ. |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Multimedia Fusion, Content Processing, Multimedia Retrieval, Digital Watermarking, and Related Topics |
| Chair | Yoichi Haneda(Univ. of Electro-Comm.) / Akinori Ito(Tohoku Univ.) |
| Vice Chair | Yukio Iwaya(Tohoku Gakuin Univ.) / Mitsunori Mizumachi(Kyushu Inst. of Tech.) / Masashi Unoki(JAIST) / Masaki Kawamura(Yamaguchi Univ.) |
| Secretary | Yukio Iwaya(NTT) / Mitsunori Mizumachi(KDDI R&D Labs.) / Masashi Unoki(Univ. of Electro-Comm.) / Masaki Kawamura(Nagasaki Univ.) |
| Assistant | Shoichi Koyama(Univ. of Tokyo) / Motoi Iwata(Osaka Pref. Univ.) / Kazuhiro Kohno(Kansai Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Engineering Acoustics / Technical Committee on Enriched MultiMedia |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Intelligibility prediction method for outdoor public address systems using classifiers |
| Sub Title (in English) | |
| Keyword(1) | outdoor public address system |
| Keyword(2) | speech intelligibility |
| Keyword(3) | intelligibility prediction |
| Keyword(4) | machine learning |
| Keyword(5) | SVM |
| 1st Author's Name | Yosuke Kobayashi |
| 1st Author's Affiliation | National Institute of Technology, Miyakonojo College(NIT Miyakonojo) |
| 2nd Author's Name | Kazuhiro Kondo |
| 2nd Author's Affiliation | Yamagata University(Yamagata Univ.) |
| Date | 2015-11-12 |
| Paper # | EA2015-26,EMM2015-47 |
| Volume (vol) | vol.115 |
| Number (no) | EA-302,EMM-303 |
| Page | pp.pp.13-18(EA), pp.13-18(EMM), |
| #Pages | 6 |
| Date of Issue | 2015-11-05 (EA, EMM) |