Daffern Helena, Keren-Portnoy Tamar, DePaolis Rory A, Brown Kenneth I
AudioLab, Department of Electronic Engineering, University of York, United Kingdom.
Department of Language and Linguistic Science, University of York, United Kingdom.
Appl Acoust. 2020 May;162:107183. doi: 10.1016/j.apacoust.2019.107183.
This project set out to develop an app for infants under one year of age that responds in real time to language-like infant utterances with attractive images on an iPad screen. Language-like vocalisations were defined as voiced utterances which were not high pitched squeals, nor shouts. The app, BabblePlay, was intended for use in psycholinguistic research to investigate the possible causal relationship between early canonical babble and early onset of word production. It is also designed for a clinical setting, (1) to illustrate the importance of feedback as a way to encourage infant vocalisations, and (2) to provide consonant production practice for infant populations that do not vocalise enough or who vocalise in an atypical way, specifically, autistic infants (once they have begun to produce consonants). This paper describes the development and testing of BabblePlay, which responds to an infant's vocalisations with colourful moving shapes on the screen that are analogous to some features of the infant's vocalization including loudness and duration. Validation testing showed high correlation between the app and two human judges in identifying vocalisations in 200 min of BabblePlay recordings, and a feasibility study conducted with 60 infants indicates that they can learn the contingency between their vocalisations and the appearance of shapes on the screen in one five minute BabblePlay session. BabblePlay meets the specification of being a simple and easy-to-use app. It has been shown to be a promising tool for research on infant language development that could lead to its use in home and professional environments to demonstrate the importance of immediate reward for vocal utterances to increase vocalisations in infants.
本项目旨在开发一款针对一岁以下婴儿的应用程序,该程序能在iPad屏幕上以吸引人的图像实时回应类似语言的婴儿发声。类似语言的发声被定义为非高音尖叫或呼喊的有声发声。这款名为“BabblePlay”的应用程序旨在用于心理语言学研究,以探究早期规范咿呀学语与单词产生的早期 onset 之间可能的因果关系。它还专为临床环境设计,(1)说明反馈作为鼓励婴儿发声方式的重要性,(2)为发声不足或以非典型方式发声的婴儿群体,特别是自闭症婴儿(一旦他们开始发出辅音)提供辅音发音练习。本文描述了“BabblePlay”的开发和测试过程,该应用程序通过屏幕上色彩斑斓的移动形状来回应婴儿的发声,这些形状类似于婴儿发声的一些特征,包括响度和时长。验证测试表明,在200分钟的“BabblePlay”录音中,该应用程序与两名人类评判员在识别发声方面具有高度相关性,并且对60名婴儿进行的可行性研究表明,他们可以在一次五分钟的“BabblePlay”会话中学习到自己的发声与屏幕上形状出现之间的偶然性。“BabblePlay”符合简单易用应用程序的规格。它已被证明是研究婴儿语言发展的一个有前途的工具,可能会导致其在家庭和专业环境中被用于证明即时奖励对发声的重要性,以增加婴儿的发声。