AI Institute for Transforming Education for Children with Speech and Language Processing Challenges

人工智能研究所致力于改变面临语音和语言处理挑战的儿童的教育

• 批准号: 2229873
• 负责人: Venugopal Govindaraju
• 金额: $ 2000万
• 依托单位: SUNY at Buffalo
• 依托单位国家: 美国
• 项目类别: Cooperative Agreement
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-01-15 至 2027-12-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2229873&HistoricalAwards=false
• 关键词: AI; Institute; Transforming; Education; Children

It is estimated that more than 3.4 million children need speech and language related services in the US school system, yet there are less than sixty-one thousand speech-language pathologists (SLPs) to serve them. The COVID-19 pandemic has further exacerbated this gap, making it almost impossible for SLPs to provide individualized services for children. The AI Institute for Transforming Education for Children with Speech and Language Processing Challenges aims to close this gap by developing advanced AI technologies to scale SLPs' availability and services such that no child in need of speech and language services is left behind. Towards this end, the Institute proposes to develop two novel AI solutions: (1) the AI Screener to enable universal early screening for all children, and (2) the AI Orchestrator to work with SLPs to provide individualized interventions for children with their formal Individualized Educational Plan (IEP). In developing these solutions, the Institute will advance foundational AI technologies, enhance understanding of children's speech and language development, serve as a nexus point for special education stakeholders, and represent a fundamental paradigm shift in how SLPs serve children in need of ability based speech and language services.The AI Screener will be initially deployed in early childhood classrooms and will analyze video and audio streams of children's classroom interactions, derive conventional speech and language measures used by SLPs, and assess novel and hard to obtain automaticity measures. The AI Orchestrator is a superset of the AI Screener with its main application in the public school classrooms. It will help SLPs to administer a wide range of evidence-based interventions and assess their effects on meeting children's individual IEP learning targets. At the core of the Orchestrator is a robust multi-agent reinforcement learning framework that can evaluate the potential benefits of different intervention practices and recommend those most appropriate for each child. Both solutions will push significant advances in self-supervised learning to address sparse and noisy data issues, multimodality perception, learning material rewriting and enrichment, and edge AI for real time processing. The Institute will develop human centered AI design methodologies to embody the solutions in a form appropriate for children’s learning. Education research and the learning sciences will inform the initial prototyping and validation, and will derive valuable insights from the field deployed solutions.The National Center for Special Education Research at the Institute of Education Sciences of the US Department of Education is partnering with NSF to provide funding for the Institute.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

据估计，美国学校系统中有340多万儿童需要言语和语言相关服务，但为他们提供服务的言语语言病理学家(SLP)不到6.1万人。新冠肺炎疫情进一步加剧了这一差距，使SLP几乎不可能为儿童提供个性化服务。有语音和语言处理挑战的人工智能教育研究所旨在通过开发先进的人工智能技术来扩大SLP的可用性和服务，以便不让任何需要语音和语言服务的儿童掉队，从而缩小这一差距。为此，研究所提议开发两个新的人工智能解决方案：(1)人工智能筛选器，以实现对所有儿童的普遍早期筛查；(2)人工智能协调器，与SLP合作，为儿童提供具有正式个性化教育计划(IEP)的个性化干预。在开发这些解决方案时，该研究所将推进基础人工智能技术，加强对儿童言语和语言发展的理解，作为特殊教育利益攸关方的连接点，并代表着SLP为需要基于能力的言语和语言服务的儿童提供服务的根本范式转变。AI Screener将首先部署在幼儿课堂上，将分析儿童课堂互动的视频和音频流，得出SLP使用的传统语音和语言测量，并评估新奇和难以获得的自动化测量。AI Orchestrator是AI Screener的超集，主要应用于公立学校的教室。它将帮助SLP实施广泛的循证干预措施，并评估其对实现儿童个别IEP学习目标的影响。Orchestrator的核心是一个强大的多智能体强化学习框架，可以评估不同干预实践的潜在好处，并推荐最适合每个儿童的干预实践。这两个解决方案都将推动在自我监督学习方面取得重大进展，以解决稀疏和噪声数据问题、多模式感知、学习材料重写和丰富，以及用于实时处理的边缘人工智能。该研究所将开发以人为中心的人工智能设计方法，以适合儿童学习的形式体现解决方案。教育研究和学习科学将为最初的原型和验证提供信息，并将从现场部署的解决方案中获得有价值的见解。美国教育部教育科学研究所的国家特殊教育研究中心正在与NSF合作，为该研究所提供资金。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Text2Video-Zero: Text-to-Image Diffusion Models are Zero-Shot Video Generators

• DOI: 10.1109/iccv51070.2023.01462
• 发表时间: 2023-03
• 期刊: 2023 IEEE/CVF International Conference on Computer Vision (ICCV)
• 作者: Levon Khachatryan;A. Movsisyan;Vahram Tadevosyan;Roberto Henschel;Zhangyang Wang;Shant Navasardyan;Humphrey Shi