| Presentation | 2016-08-09 Estimation of forced-selection speech intelligibility by comparing objective measures between candidate words Kazuhiro Kondo, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | We proposed an estimation method for forced pairwise selection speech intelligibility test that takes into account the forced selection manner of the test. The distance measure used was based on the frequency-weighted segmental SNR (fwSNRseg) in sub-bands defined in the Articulation Index (AI) standard. We compared the distance between the degraded test word, and the clean templates for the two candidate words in the word pair, in each of the AI bands. The ratio of sub-bands with higher SNRseg for the correct word vs. the total number of bands, with weighting defined in the AI standard to account for the sensitivity of the human auditory system, is calculated to quantify the match between the test and the correct word characteristics. A logistic mapping function from this ratio to intelligibility using speech mixed with known noise was trained, and then used to estimate the intelligibility of speech mixed with unknown noise. The correlation and Root Mean Square Error (RMSE) between subjective and estimated intelligibility were about 0.97 and 0.10, respectively. This level of estimation accuracy should enable screening of subjective intelligibility testing conditions, minimizing the need for actual test using human subjects. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | speech intelligibility / objective measures / two-to-one forced selection / DRT / speech quality estimation |
| Paper # | EA2016-20 |
| Date of Issue | 2016-08-02 (EA) |
| Conference Information | |
| Committee | EA / ASJ-H |
|---|---|
| Conference Date | 2016/8/9(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Tohoku Gakuin Univ., Tagajo Campus |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Ultra-Realistic Acoustics, Engineering/Electro Acoustics, Psychological and Physiological Acoustics, and Related Topics |
| Chair | Mitsunori Mizumachi(Kyushu Inst. of Tech.) / Tatsuya Hirahara(富山県立大) |
| Vice Chair | Yoichi Haneda(Univ. of Electro-Comm.) / Suehiro Shimauchi(NTT) / Shigeto Furukawa(NTT) |
| Secretary | Yoichi Haneda(KDDI R&D Labs.) / Suehiro Shimauchi(Akita Pref. Univ.) / Shigeto Furukawa(リオン) |
| Assistant | Shigeto Takeoka(Shizuoka Inst. of Science and Tech.) / TREVINO Jorge(Tohoku Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Engineering Acoustics / Auditory Research Meeting |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Estimation of forced-selection speech intelligibility by comparing objective measures between candidate words |
| Sub Title (in English) | |
| Keyword(1) | speech intelligibility |
| Keyword(2) | objective measures |
| Keyword(3) | two-to-one forced selection |
| Keyword(4) | DRT |
| Keyword(5) | speech quality estimation |
| 1st Author's Name | Kazuhiro Kondo |
| 1st Author's Affiliation | Yamagata University(Yamagata Univ.) |
| Date | 2016-08-09 |
| Paper # | EA2016-20 |
| Volume (vol) | vol.116 |
| Number (no) | EA-179 |
| Page | pp.pp.1-6(EA), |
| #Pages | 6 |
| Date of Issue | 2016-08-02 (EA) |