San Diego, CA, August 4, 2010 -- Now in its fifth year, the Temporal Dynamics of Learning Center (TDLC) is looking forward to what it hopes will be a second lease on life if the National Science Foundation (NSF) renews funding for five more years, at $4 million per year. A NSF review board recently visited TDLC’s home base at the University of California, San Diego, to meet with researchers from all 10 participating universities and determine the future of the Center.
TDLC is a Science of Learning Center, one of six funded by NSF. The center’s purpose is to understand the role of time and timing in learning, and how to apply this understanding to improve the way students are educated.
|TDLC Executive Director, Principal Investigator and computer science professor Garrison Cottrell presented opening remarks on the first day of the NSF review, stating that "the goal of the TDLC is to study how timing affects learning from the synapse to the classroom."|
Across its member campuses, the TDLC is comprised of 40 principal investigators and 50 projects. Three TDLC researchers are based at the UC San Diego division of the California Institute for Telecommunications and Information Technology (Calit2), and work on a number of projects takes place at Calit2, including the Center’s robot projects and research in expression recognition. Calit2 also hosted the demo and poster session during the site visit.
TDLC’s research addresses four key questions: 1) How is temporal information about the world learned and how does the temporal dynamics of the world influence learning? 2) What are the temporal dynamics of brain cells, brain systems and behavior, and how do these cells change with learning? 3) What are the temporal dynamics of motor control and information seeking, and how are they learned? and 4) What are the general principles that explain the dynamics of learning across multiple scales and domains?
“The NSF review was an opportunity to demonstrate that we are more than the sum of our parts,” remarked Marian Stewart Bartlett, an associate research professor at UCSD’s Institute for Neural Computation and co-director of UCSD’s Machine Perception (MP) Laboratory (MPLab), based at Calit2. “Many important collaborations have come about as a result of this center, and one thing that has been measured is that major scientific breakthroughs most often take place in multidisciplinary groups. It is always important to connect people who may not have been connected before."
|CERT, or Computer Expression Recognition Toolbox, is a software package for automatically detecting facial expressions. It's a signature MPLab project that has fostered multiple collaborations, spin-off projects and even consumer technologies.|
The panel also heard testimonials from TDLC-affiliated researchers about the importance of their collaborations, which, they say, would have been less likely were the investigators funded independently.
"The TDLC has created a network of research networks in sensorimotor learning, interactive memory systems, perceptual expertise, and social interaction," noted Cottrell., a professor of computer science and engineering in the UC San Diego Jacobs School of Engineering. "This structure organizes our scientists into smaller, interdisciplinary research communities that become the engines of TDLC’s research.
"The research structure we have works," he added. "When you share techniques, vocabulary, and intellectual problems, collaboration is easier. This is a great opportunity for our trainees, as they become familiar with researchers from many disciplines who are willing to give them guidance on their projects. As one of our students once put it, ‘It takes a network to raise a graduate student.’”
CERT, or Computer Expression Recognition Toolbox, is a signature MPLab project that has fostered multiple collaborations, spin-off projects and even consumer technologies. A software package for automatically detecting facial expressions, CERT contains "smile detection" technology that allows a “teaching robot” like RUBI to notice when a child is smiling – and therefore carry on with her lesson (if the child is frowning in confusion, RUBI might take a step back and review the lesson). This smile detection technology was created using machine learning techniques, trained upon a dataset of more than 60,000 individuals.
These smile detector algorithms are used in Sony Shutter Smile Technology, which is found in the brand's latest generation of consumer digital camera products. CERT also has been "married" with a program developed by Jim Tanaka and colleagues at the University of Victoria called "Let's Face It," which trains children with autism how to recognize facial expressions. The combined systems train the children to produce facial expressions by providing them feedback on how accurately they are producing them.
"The take-home message," said Cottrell, "is that technological advances have spread through the TDLC, spawning new research directions and potential. Our goals are being well-served by our projects."
The next step for TDLC members will be to respond to the NSF’s Site Visit Report, which was delivered in mid-July. The final determination concerning the Center's funding will come in February of 2011. If funded, TDLC plans to continue to engage in a number of collaborative projects, including research into the role of music and rhythmicity, and how they relate to brain rhythms associated with readiness for learning.
Story written by Tiffany Fox, Calit2.
Robot Learns to Smile »
Your Face as a Remote Control »
Temporal Dynamics of Learning Center »
Machine Perception Laboratory »
National Science Foundation »
More Computer Science and Engineering News Via RSS More Mechanical and Aerospace Engineering News Via RSS
Get the monthly newsletter from the Jacobs School of Engineering