SCH: INT: Conversations for Vision: Human-Computer Synergies in Prosthetic Interactions

SCH：INT：视觉对话：假肢交互中的人机协同

• 批准号: 10220137
• 负责人: John Millar Carroll
• 金额: $ 25.83万
• 依托单位: PENNSYLVANIA STATE UNIVERSITY, THE
• 依托单位国家: 美国
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-09-19 至 2024-07-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10220137
• 关键词: Address; Adult; American; Articulation; Back; Blindness; Communities; Computer Vision Systems; Computers; Data Set; Development; Economics; Emotional; Female; Goals; Human; Information Sciences; Instruction; Mediating; Modeling; Ocular Prosthesis; Performance; Prosthesis; Route; Self-Help Devices; Series; Services; Social Well-Being; Technology; Time; Underrepresented Students; Vision; Visual; Visual impairment; Work; Workplace; base; blind; design; experience; graduate student; human-in-the-loop; learning materials; legally blind; outreach; prototype; simulation; synergism; undergraduate student

The project will investigate prosthetic support for people with visual impairment (PVI) that integrates computer vision-based prosthetics with video-mediated human-in-the-loop prosthetics. Computer vision- based (CV) prosthetics construe the fundamental technical challenge for visual prosthetics as one of parsing and identifying objects across scales, distances, and orientations. Visual prosthetic applications have been central drivers in the development of computer vision technology through the past 50 years. Video-mediated remote sighted assistance (RSA) prosthetics are more recent, enabled by different technologies, and construe the orienting technical challenge for visual prosthetics as one of effective helping interactions. RSA services are commercially available now, and have evoked much excitement in the PVI community. The two approaches, CV and RSA, will be successively integrated through a series of increasingly refined Wizard of Oz simulations, and investigate possible synergies between the two approaches. We will employ a human-centered design approach, identifying a set of key assistive interaction scenarios that represent authentic needs and concerns of PVIs, by leveraging our 6-year relationship working directly with our local chapter of the National Federation of the Blind. RELEVANCE (See Instructions): 23.7 million American adults have vision loss; 1.3 million people in US are legally blind. This project addresses a transformational opportunity to enhance human performance and experience, to diversify workplace participation, and to enhance economic and social well-being.

该项目将研究视力障碍（PVI）患者的假肢支持， 基于计算机视觉的假肢与视频介导的人在环假肢。计算机视觉- 基于CV的视觉修复是视觉修复的基本技术挑战， 以及跨尺度、距离和方向识别对象。视觉假体的应用已经 在过去的50年里，计算机视觉技术发展的核心驱动力。视频介导 远程视力辅助（RSA）假肢是最近才出现的，由不同的技术实现， 视觉修复的定向技术挑战是有效的帮助互动之一。RSA服务 现在已经商业化，并且在PVI社区中引起了很大的兴奋。 CV和RSA这两种方法将通过一系列日益完善的 绿野仙踪模拟，并调查这两种方法之间可能的协同作用。我们会委聘 以人为本的设计方法，确定一组关键的辅助交互场景， 通过利用我们6年的合作关系，直接与我们当地的 全国盲人联合会的分会。 相关性（见说明）： 23.7 100万美国成年人视力丧失; 130万美国人是法定盲人。该项目解决了 转型机会，以提高人的绩效和经验，使工作场所的参与多样化， 提高经济和社会福利。