Presentation | 2017-10-20 Low Cost Semi-automatic Correction and Adaptation Method Assuming Automatic Captioning System for Lectures Tamiya Kenta, Terada Yuji, Kai Atsuhiko, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | By using Automatic Speech Recognition (ASR) technology, it is possible to subtitle lecture and other voices at low cost and in real time, which is a great help for the hearing impaired people. However, when using the ASR system, there is a problem that the recognition accuracy is greatly influenced by the fact that the technical term tends to become an unknown word especially in a university lecture and the recognition accuracy is greatly influenced by the speaker and the recording environment. In order to correct such a misrecognition result, conventional semi-automatic captioning systems require several operators for simultaneous editing, or cause a large delay for time-consuming editing work. In this paper, we propose a low cost correction method to feedback only a part of errors such as misrecognized technical terms and to identify and correct erroneously recognized segments by using Spoken Term Detection (STD) and lattice modification methods. We also adopt an unsupervised language model adaptation for additional subtitle correction after the modified online caption text were obtained for a lecture. We report the experimental result of our proposed system using the lecture speech corpus. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Automatic Speech Recognition / Spoken Term Detection / Automatic captioning system / Recognition error correction / Supporting hearing impaired |
Paper # | SP2017-50,WIT2017-46 |
Date of Issue | 2017-10-12 (SP, WIT) |
Conference Information | |
Committee | WIT / SP |
---|---|
Conference Date | 2017/10/19(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Tobata Library of Kyutech (Kitakyushu) |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | Chikamune Wada(Kyushu Inst. of Tech.) / Yoichi Yamashita(Ritsumeikan Univ.) |
Vice Chair | Daisuke Wakatsuki(Tsukuba Univ. of Tech.) / Hiroki Mori(Utsunomiya Univ.) |
Secretary | Daisuke Wakatsuki(Nagoya Inst. of Tech.) / Hiroki Mori(AIST) |
Assistant | Takeaki Shionome(*) / Manabi Miyagi(Tsukuba Univ. of Tech.) / Takashi Handa(Saitama Industrial Technology Center) / Kei Hashimoto(Nagoya Inst. of Tech.) / Satoshi Kobashikawa(NTT) |
Paper Information | |
Registration To | Technical Committee on Well-being Information Technology / Technical Committee on Speech |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Low Cost Semi-automatic Correction and Adaptation Method Assuming Automatic Captioning System for Lectures |
Sub Title (in English) | |
Keyword(1) | Automatic Speech Recognition |
Keyword(2) | Spoken Term Detection |
Keyword(3) | Automatic captioning system |
Keyword(4) | Recognition error correction |
Keyword(5) | Supporting hearing impaired |
1st Author's Name | Tamiya Kenta |
1st Author's Affiliation | Shizuoka University(Shizuoka Univ.) |
2nd Author's Name | Terada Yuji |
2nd Author's Affiliation | Shizuoka University(Shizuoka Univ.) |
3rd Author's Name | Kai Atsuhiko |
3rd Author's Affiliation | Shizuoka University(Shizuoka Univ.) |
Date | 2017-10-20 |
Paper # | SP2017-50,WIT2017-46 |
Volume (vol) | vol.117 |
Number (no) | SP-250,WIT-251 |
Page | pp.pp.89-94(SP), pp.89-94(WIT), |
#Pages | 6 |
Date of Issue | 2017-10-12 (SP, WIT) |