Basic Research for Employng Voice Recognition Systems in the Education of the Deaf and Hearing Impaired.

在聋人和听力障碍者教育中使用语音识别系统的基础研究。

• 批准号: 11680290
• 负责人: HOSOYA Miyoko
• 金额: $ 1.98万
• 依托单位: Tsukuba College of Technology
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (C)
• 财政年份: 1999
• 资助国家: 日本
• 起止时间: 1999 至 2000
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-11680290/
• 关键词: voice recognition; speech recognition; Deaf and Hearing Impaired; school for the deaf and hearing impaired; captioning; real-time; Via Voice; 音声認識; 言語教育

1999-1 : For this research we used the IBM ViaVoice speech recognition software. After a period of enrollment/training the lectures of the previous year were input from printed data, thereby extending the vocabulary and other software components to further improve the recognition accuracy.1999-2 : Assessed the use of headsets during lectures. While the students had no difficulty lip reading, the lecturer had a little difficulty using sign language. Consequently, it was decided that a cordless microphone was preferable.1999-3 : Erroneous display of kanji put the students at a great disadvantage. This disadvantage was overcome by using only kana without kanji.2000-1 : Tests were set up outside college at a kindergarten and a high school for the Deaf and Hearing Impaired. The information obtained from the teachers and tutors proved invaluable.2000-2 : During individual tutoring outside regular lecture time, correction of miss-recognition was easy and when compared to conversing in writing, it was evident that the system was very effective.2000-3 : Analyzed reports from an Australian school for the deaf and hearing impaired. The results indicated the importance of conducting comparative investigations of a number of voice recognition software packages.

1999-1：在这项研究中，我们使用了IBM ViaVoice语音识别软件。经过一段时间的注册/培训，前一年的讲座是从打印数据输入，从而扩展词汇和其他软件组件，以进一步提高识别精度。1999 -2：评估在讲座期间使用耳机。虽然学生们在唇阅读方面没有困难，但讲师在使用手语方面有一点困难。因此，决定使用无绳麦克风。1999 -3：错误的汉字显示使学生处于非常不利的地位。2000 -1：在大学外的一所幼儿园和一所聋哑人高中设立了测试。从教师和导师获得的信息证明是宝贵的。2000 -2：在个别辅导定期讲座时间以外，纠正错误识别是很容易的，当与书面交谈，这是很明显的系统是非常有效的。2000 -3：分析报告，从澳大利亚的聋人和听力受损的学校。研究结果表明，对若干语音识别软件包进行比较研究十分重要。

项目成果

小林正幸: "聴覚障害者のための音声認識を活用したリアルタイム字幕挿入システム(2)"電子情報通信学会技術研究報告. 100. 129-134 (2001)

Masayuki Kobayashi：“针对听力障碍者的使用语音识别的实时字幕插入系统（2）”IEICE 技术研究报告。100. 129-134（2001）。

石原保志: "大学等の講義における聴覚障害を対象とした情報保障の方法論的検討-手話通訳・リアルタイム文字呈示・要約解説の比較-"電子情報通信学会技術研究報告. 100. 7-14 (2001)

石原靖：“大学讲座中听障人士信息安全的方法论研究——手语解释、实时文本呈现和摘要解释的比较”IEICE技术研究报告100. 7-14 (2001)。

KOBAYASHI Masayuki: "Real-Time Captioning System Using Voice Recognition for Hearing Impaired Person (1)"TECHNICAL REPORT of EICE. ET99-89. 41-48 (2000)

小林雅之：“针对听力障碍人士的使用语音识别的实时字幕系统（1）”EICE 的技术报告。

石原保志: "青年期の発音指導"聴覚障害. 54. 4-9 (1999)

Yasushi Ishihara：“青少年发音指导”听力障碍。 54. 4-9 (1999)