Computer Science Professor Wins NSF CAREER Grant
A newly awarded National Science Foundation grant will enable a University of Arizona computer science faculty member to improve applications of computer imaging to everything from designing robot assistants to interpreting data from drug screens.
Kobus Barnard, an assistant professor in the computer science department, has been awarded a $450,000 Faculty Early Career Development Program, or CAREER, grant.
â€œTo help us deal with the world, computers and robots need to know where objects begin and end, what they are â€“ is it a table or a chair â€“ and what they can be used for," Barnard said. â€œThese are difficult computational problems that we are addressing by learning statistical geometric models for objects from large data sets."
The grant goes to faculty members who exemplify the role of teacher-scholars through outstanding research, excellent education and the integration of education and research within the context of the mission of their organizations. Barnardâ€™s grant will span five years and will focus on research, education and K-12 outreach.
â€œCAREER grants are NSFâ€™s most prestigious awards for young faculty. Consequently, they are very hard to get,â€ said Greg Andrews, interim department head of computer science. â€œThis award to Kobus is a tribute to his accomplishments and promise.â€
Barnardâ€™s research focuses on understanding and recognizing objects in digital images, from everyday photos to images captured for biological and biomedical research.
Barnard and his colleagues develop 3-D geometric models of objects to match digital image data, thereby recognizing objects, such as tables and chairs, in images and determining what size the objects are and how they are arranged in space.
For example, an intelligent robotic system for assisted living would need to recognize whether a table is blocking a path to the door. Being able to recognize objectsâ€™ form, dimensions and location in space is critically important for many problems in computer vision, intelligent systems and computational biology.
Matching geometric models to digital images of biological structures, ranging from crop plants to fungal pathogens to brain neurons, allows researchers to study the connections between shape and function.
The techniques Barnard and his colleagues are developing will enable scientists to determine the links between an organismâ€™s detailed shape and the organismâ€™s molecular makeup and genes as well as its environment. The result is that scientists will be able to develop new ways to automate data analysis for important activities such as breeding crops, screening for new pharmaceuticals and recognizing harmful species.
Barnardâ€™s project also includes working with the UAâ€™s College of Education and several other departments.
He will develop curriculum for UA classes in computer vision, machine learning and scientific applications and also collaborate with Sara Chavarria, the College of Educationâ€™s director of education and outreach, to design and implement an educational outreach program. The program will target Tucson high school students from low socioeconomic groups and will promote an understanding of the integration of science and computation.
In applying his work to botanical and biomedical problems, Barnard also will collaborate with faculty members in several UA departments, including Barry Pryor, associate professor of plant pathology; Linda Restifo, professor of neurobiology and of neurology; and Ravi Palanivelu, assistant professor of plant science. Students involved in the project include UA graduate student Joseph Schlecht and UA computer science graduate Kate Spriggs.