Sensorimotor learning of facial expressions: A novel intervention for autism

描述（由申请人提供）：面部表情的感觉运动学习：自闭症的一种新的干预措施项目摘要：本申请涉及广泛的挑战领域（04）：临床研究，以及具体的挑战主题04- mh -101自闭症：应对挑战：新的干预措施。患有自闭症谱系障碍（ASD）的儿童产生和感知动态面部表情的能力受损（Adolphs, Sears, and Piven, 2001）。先进的计算机视觉技术现在可以用于研究自闭症谱系障碍（ASD）儿童常见的面部表情识别和生产缺陷等问题。这些技术不仅可以帮助量化这些赤字，而且还可以用作旨在降低赤字严重程度的干预措施的一部分。在这个项目中，自动面部表情识别将被用于开发自闭症儿童面部表情生成的训练练习。Tanaka和Schultz之前的研究开发了Let's Face It！干预工具可有效改善ASD儿童的面部处理能力。让我们面对现实吧！专注于感知，而不是生产。Bartlett和Movellan开发了计算机表情识别工具箱（CERT），这是一种计算机视觉系统，可以实时测量37个面部表情维度。该项目将这两个研究小组联合起来，以开发一种计算机辅助干预系统，以提高自闭症谱系障碍（ASD）儿童的社交技能。计算机视觉工具使专注于面部表情产生的新颖干预成为可能。干预游戏将被开发出来，以一种吸引人且经济有效的方式，为自闭症儿童的面部表情提供自动反馈。除了提高生产技能外，面部表情的生产练习也可以通过将运动生产与感知联系起来来提高感知技能。这里开发的干预练习将包括闭环感觉运动表达训练，在这种训练中，受试者看到并产生面部表情，然后观察他们的面部表情在游戏中的效果。该项目还将在一系列评估任务中表征ASD儿童的面部表情产生，包括使用CERT自动面部表情测量。完成后，该项目将有助于我们了解ASD儿童的面部表情产生，为面部表情产生提供新的干预工具，并提供干预效果的数据。本文提出的干预系统可能对任何非典型面部表情产生患者群体的社会发展产生巨大影响，包括ASD和注意缺陷障碍。该项目还将有助于了解对社会功能至关重要的运动技能的发展和学习。自动面部表情识别技术为临床研究、评估和干预系统开辟了新的可能性。将此类技术整合到临床研究中既及时又至关重要。本文提出的干预系统可能对任何非典型面部表情产生患者群体的社会发展产生巨大影响，包括ASD和注意缺陷障碍。该项目还将有助于了解对社会功能至关重要的运动技能的发展和学习。

DESCRIPTION (provided by applicant): Sensorimotor learning of facial expressions: A novel intervention for autism Project Summary This application addresses broad Challenge Area (04): Clinical Research, and specific Challenge Topic 04-MH-101 Autism: Addressing the challenge: Novel interventions. Children with Autism Spectrum Disorders (ASD) are impaired in their ability to produce and perceive dynamic facial expressions (Adolphs, Sears, and Piven, 2001). Advanced computer vision technologies can now be leveraged in the investigation of issues such as the facial expression recognition and production deficits common to children with autism spectrum disorder (ASD). Not only can these technologies assist in quantifying these deficits, but they can also be used as part of interventions aimed at reducing deficit severity. In this project, automated facial expression recognition will be employed for the development of training exercises for facial expression production in children with ASD. Previous research by Tanaka and Schultz developed the Let's Face It! intervention tool shown to effectively improve the face processing abilities of children with ASD. Let's Face It! focused on perception but not production. Bartlett and Movellan have developed the Computer Expression Recognition Toolbox (CERT), a computer vision system which measures 37 facial expression dimensions in real-time. This project brings together these two research groups in order to develop a computer assisted intervention system to enhance the social communication skills of children with Autism Spectrum Disorder (ASD). The computer vision tool enables novel interventions that focus on facial expression production. Intervention games will be developed to provide automated feedback on facial expressions produced by children with ASD in an engaging and cost- effective manner. In addition to improving production skills, practice with facial expression production may also improve perception skills by linking motor production with perception. The intervention exercises developed here will include closed-loop sensorimotor expression training in which subjects both see and produce facial expressions, and then view the effects of their facial expressions in the games. This project will also characterize facial expression production of children with ASD in a battery of assessment tasks, including automated facial expression measurement with CERT. Upon completion, this project will contribute to our understanding of facial expression production in children with ASD, provide a new intervention tool for facial expression production, and provide data on the efficacy of the intervention. The intervention system proposed here could have tremendous impact on social development in any patient population with atypical facial expression production, including ASD, and attention deficit disorder. The project will also contribute to the understanding of the development and learning of motor skills essential to social functioning. Sensorimotor learning of facial expressions: A novel Intervention for autism Project Narrative Automated facial expression recognition technology opens new possibilities for clinical research, assessment, and intervention systems. Integration of such technology in clinical research is both timely and crucial. The intervention system proposed here could have tremendous impact on social development in any patient population with atypical facial expression production, including ASD, and attention deficit disorder. The project will also contribute to the understanding of the development and learning of motor skills essential to social functioning.