MRI Development: Enabling Research in Natural Communication with Virtual Tutors, Therapists, and Robotic Companions

该项目开发了SocioBot-SDS（SocioBot-Spoken Dialog System），这是一种具有情感反应的机器人角色形式的工具，预计将推进涉及下一代人机交互的研究。机器人仪器将用于治疗和教育目的。它的发展将特别有助于感知言语和视觉行为的研究领域。该仪器的组件集成了独特的可编程性和鲁棒性水平，以显示人类和非人类的情感姿态，并通过关节颈进行传统的方向控制。这项工作预计将加速社交机器人的研究和开发，这些机器人可以准确地模拟与敏感而有效的人类导师，临床医生或护理人员进行面对面交流的动态，达到目前仪器无法实现的程度。机器人代理建立在计算机视觉，口语对话系统，角色动画和有效计算的基础上，进行对话，与用户建立融洽的关系，产生丰富的情感面部表情，与韵律语音生成同步，以响应用户的语音和情感。该仪器代表了情感能力集成的新水平，使研究人员能够研究社交情感/机器人/代理，这些机器人/代理可以理解口语并表达情感，并以自然的方式与人进行有效的互动，说话和沟通（就像人类一样）。该仪器为支持跨学科技术的研究和培训提供了令人兴奋的平台。由于研究界可以使用该仪器，因此可以通过感知交互者的语音模式和视觉行为来加速研究如何优化人与化身之间的沟通。该仪器代表了情感能力整合的新水平，并作为设计新一代更具沉浸式和有效的智能辅导和治疗系统的平台，以及针对人类残疾的机器人辅助治疗，包括有感官，注意力和语言延迟风险的婴儿以及患有精神残疾的成年人。 从教育的角度来看，拟议的活动将通过让各级学生在文书制定期间和之后参与，加强跨学科教育。由此产生的仪器将免费分发给研究人员，以研究机器人行为，这些行为导致各种任务领域的沉浸式和有效的应用，例如教学生阅读，辅导学生学习科学，进行言语和语言治疗课程，或为家中的老年人提供陪伴。此外，这一发展启动的活动加强了跨学科教育，涉及本科生和研究生一级的学生，在仪器的开发期间和之后。

This project, developing SocioBot-SDS (SocioBot-Spoken Dialog System), an instrument in the form of a robotic character with an emotional response, is expected to advance research involving next generation human-machine interactions. The robotic instrument will be used for therapeutic and educational purposes. Its development will specifically contribute to the research area of perceived speech and visual behaviors. The components of the instrument integrate a unique level of programmability and robustness to the display of human- and non-human-like emotive gestures with conventional orientation control through an articulated neck. The work is expected to accelerate research and development of social robots that can accurately model the dynamics of face-to-face communication with a sensitive and effective human tutor, clinician, or caregiver to a degree unachievable with current instrumentation. The robotic agent builds on advances in computer vision, spoken dialogue systems, character animation and effective computing to conduct dialogues that establish rapport with users producing rich, emotive facial gestures synchronized with prosodic speech generations in response to users' speech and emotions. The instrument represents a new level of integration of emotive capabilities that enable researchers to study socially emotive/robots/agents that can understand spoken language and show emotions and interact, speak, and communicate effectively with people in a natural way (as humans do).The instrument provides an exciting platform for research and training supporting cross-discipline technology. Since the research community would have access to the instrument, research in how to optimize communication among people and avatars might be accelerated through the perception of speech patterns and visual behaviors of those interacting. The instrument represents a new level of integration of emotive capabilities and serves as a platform for designing a new generation of more immersive and effective intelligent tutoring and therapy systems, and robot-assisted therapeutic treatments for human disabilities that include infants at risk for sensory, attention and language delays, as well as adults with mental disabilities. From an educational perspective, the proposed activities will enhance inter-disciplinary education by involving students at all levels, during and beyond the development of the instrument. The resulting instrument will be freely distributed for researchers to investigate robotic behaviors that lead to immersive and effective applications across a variety of task domains such as teaching students to read, tutoring students in science, conducting speech and language therapy sessions, or providing companionship to elderly individuals in their homes. Moreover, the activities initiated by this development enhance interdisciplinary education, involving students at the undergraduate and graduate levels, during and beyond the development of the instrument.