...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Proceedings of the National Academy of Sciences of the United States of America >Fully automated method to study early language development
【24h】

Fully automated method to study early language development

机译：全自动方法研究早期语言发展

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Research in early language development is currently restricted by laborious human analysis of small recorded samples, limiting the scope of study and the potential to detect emerging disorders. As a proof of concept for fully automated language analysis, D. Kimbrough Oiler et al. (pp. 13354-13359) collected nearly 1,500 all-day soundtracks from over 200 children, aged I 10 months to 4 years, with battery-powered recorders worn in a chest pocket of the children's clothing. An automated system separated sounds made by the child from sounds in the environment and then rated the child's utterances according to characteristics defined by vocal development theory. Reliability tests on 70 one-hour sample recordings revealed that the algorithm recognized specific vocal elements in agreement with human assessments approximately 70% of the time, with a low false-positive rate.
机译：当前，对早期语言发展的研究受到人类对少量记录样本进行费力的分析的限制，从而限制了研究范围和发现新出现疾病的可能性。作为全自动语言分析的概念证明，D。Kimbrough Oiler等人。 （pp。13354-13359）从200多名年龄在10个月至4岁的200名儿童中收集了近1,500条全天配乐，电池供电的录音机戴在儿童衣服的胸前口袋中。一个自动系统将孩子发出的声音与环境中的声音分开，然后根据语音发展理论定义的特征对孩子的话语进行评分。对70个1小时样本记录的可靠性测试表明，该算法大约70％的时间与人类评估一致地识别了特定的声音元素，而且假阳性率很低。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号