Today, the U.S. National Science Foundation launched a shared research infrastructure that will provide artificial intelligence researchers and students across the nation with access to innovative resources containing high-quality data on human-machine interactions in the context of collaborative teams. announced a $16.1 million investment to support Autonomous Driving and News Recommendations. The project includes a platform for conducting AI research on social robotics and immersive virtual environments.
The award is part of NSF’s Computer and Information Science and Engineering Community Research Infrastructure (CCRI) program, which aims to create, enhance, and democratize access to research infrastructure, and is designed to help the computer and information sciences. and engineering.
“A key component of the success of the AI research revolution is enabling researchers to access the data and platforms they need to continue driving innovation and scalability in AI technologies and systems,” said NSF Director Sethuraman Panchanathan. says Mr. “This infrastructure must be accessible to a wide and diverse pool of people interested in AI research and development, as it drives modern discovery.”
The investment will support five collaborative projects led by the University of Central Florida. University of Pennsylvania. University of Minnesota Twin Cities. UCLA; and Pennsylvania. Collectively, the institutions provide hands-on training and educational opportunities for researchers and students, and create communities that enable knowledge sharing and collaborative tools to strengthen and strengthen the nation’s AI-advanced workforce. Zoom in.
The award will focus on the following aspects of AI research infrastructure and equipment:
An open-source simulation platform for AI research on autonomous driving
Researchers in this UCLA-led project aim to develop an open-ended driving simulation platform that will foster innovation in many aspects of autonomous driving. The platform supports realistic driving simulations using various traffic assets and scenarios imported from the real world. It is envisioned as a common testing ground for researchers in academia and industry to develop new AI methods, share data and models, and benchmark progress.
Autonomous driving is expected to transform everyday life and the economy, promising benefits such as safe transportation and efficient mobility. However, much of today’s research into autonomous driving is done in expensive commercial vehicles. This is an expensive and dangerous way to evaluate AI and machine learning methods. The team’s proposed driving simulator infrastructure provides a cost-effective and safe alternative for developing and evaluating new AI algorithms for autonomous driving. This research is conducted in collaboration with the University of California, Berkeley.
Virtual Experience Research Accelerator (VERA)
Led by the University of Central Florida, this project focuses on sharing resources for conducting research in Augmented Reality (XR) (virtual, augmented and mixed reality environments). The team will combine and expand aspects of distributed lab-based research, online research, research panels, and crowdsourcing to support researchers in the field of lab-based human subjects research, transforming it for the research community. We aim to develop a comprehensive infrastructure.
VERA will provide the data and scalability the XR community needs, and promote equity by providing researchers (including students) across the country with the opportunity to conduct high-impact research, even if they do not have access to XR. is designed to Locate labs and instruments locally. VERA also supports the creation of large and diverse training data sets for AI related to XR. The research will be conducted in collaboration with Cornell Institute of Technology, Davidson University, Lehigh University, and Stanford University.
Integrate, grow and sustain research communities around a modular social robotics platform
Led by the University of Pennsylvania, researchers in this project aim to build a standardized infrastructure to create a collaborative platform for robotics and AI research. The project will build 50 humanoid robots and distribute them to selected research teams across the United States to enable the development of novel approaches to online data collection, collaboration, and AI decision-making.
Additionally, the team seeks to provide training for researchers to work collaboratively and create a community of roboticists where they can learn from each other and share ideas. The expected result is rapid progress in robotics and human-robot interaction research in the United States and an increase in the diversity of new teams of robotics researchers. This research is conducted in collaboration with Oregon State University.
Live User Research News Recommender Infrastructure for Algorithm and Interface Experimentation
Led by the University of Minnesota Twin Cities, the project will bring researchers across the country to study live, one-time, long-term interactions between users and AI systems that personalize user experiences based on past behavior. We aim to develop a shared news recommendation system that will allow us to do so. The project team includes skilled experimental researchers who have conducted extensive research into end-user privacy, including research participant privacy, and ethics experts who address the privacy-related implications of these technologies for end-users. will be
Recommender systems have a very wide range of effects, such as ranking and displaying products to online shoppers based on their past shopping behavior. Recommender systems are also behind most online news sources and can shape the news people see. Given the importance of these systems, it is important that researchers are able to conduct studies to evaluate different algorithm and interface designs and their impact on users. The project will be conducted in collaboration with Clemson University, Boise State University, Northwestern University, and the University of Colorado Boulder.
An open data infrastructure for understanding embodied emotions
Led by Pennsylvania State University, this project aims to unlock the wealth of information about human representation found in videos on the Internet. Similar to other areas of AI such as image recognition, human body movements (including emotions) can provide essential insights for the development of future human-machine interaction systems. Sentiment analysis on text has received a great deal of attention, but the analysis of sentiment and sentiment from non-textual inputs has been less explored and could be greatly advanced by the infrastructure developed in this project. I have.
Our multidisciplinary research team is made up of experts in fields such as AI, computer vision, expressive robotics, emotion recognition, psychology, statistics, and data mining. They work with research ethics experts to ensure that considerations regarding the use of images in AI training and their application in projects follow best practices. This research will be conducted in collaboration with the University of Illinois at Chicago and the Robotics, Automation and Dance Lab.
