社會情緒互惠(Social-emotional reciprocity)缺損為自閉症類群障礙症(Autism Spectrum Disorder, ASD)主要障礙之一,特別是基本臉部表情表達與推論情緒能力缺乏,此障礙影響此族群於日常生活中與他人正常社交互動行為。擴增實境技術(Augmented reality, AR)已被許多研究證實對自閉症類群障礙症之好處,如抽象概念視覺化、虛實結合與即時互動等。但目前大部分研究對此族群仍缺乏探討情緒表達與推論情緒之能力探究。因此,本研究針對此族群之臉部表情表達與推論情緒障礙開發擴增實境情緒訓練與推理系統(AR-Emotion system),其系統教學模式採用基礎及複雜臉部情緒辨識到社會情境情緒推論,幫助ASD孩童學習辨識八種情緒與情緒推論能力;系統功能包含AR圖卡感應互動模式,3D臉部動畫及社會情境,及文字、語音、提示及圖卡即時反饋等功能。因此本研究探究AR-Emotion系統對ASD孩童的社會情緒辨識與情緒推論技能之學習成效。本研究方法使用跨受試之多基線設計(Cross-subjects multiple baseline design),透過視覺分析結果觀察目標行為變化。研究對象為三名ASD孩童在12-13週內進行實驗。結果顯示,在使用AR-Emotion系統後,每位參與者的情緒辨識與情緒推論技能表現進步狀況,且本系統可吸引自閉症孩童注意力,提升孩童學習動機與練習機會。因此,本系統對於教導情緒辨識與推論能力有良好之學習成效。
Social-emotional reciprocity is one of the major social deficits for autism spectrum disorder (ASD), especially the lack of capabilities of expressing and facial expressions. This deficit deeply affects their social interaction in daily life. Augmented reality (AR) technology has been proven to be beneficial for children with ASD ; for example, providing visualizing information, and combining the real-word and real-time interaction. However, most research still lacks of using AR technology in exploring facial expression and inference in this population. Therefore, the purpose of this study was to develop an AR Emotion System, the teaching content of which involved recognizing and inferring basic and complex emotions for this population. The functions of the system included mark-based cards for interacting with 3D animated models and 3D animated social context, which also involved speaking sounds, text and prompts to help children with ASD to enhance their capabilities of recognizing and inferring emotions. Therefore, this study investigated the learning effects of using the AR-Emotion system for recognizing and inferring social emotions in children with ASD.A cross-subjects multiple baseline design and visual analysis were used for observing targeted behaviors of these participants. Three children with ASD were engaged within the experiment of 12-13 weeks for baseline, intervention and maintenance sessions. The results showed that the participants performed well in recognizing and inferring social emotions after using the AR-Emotion System.