This summer, Cornell’s outstanding contributions to the field of computer vision were on full display at the IEEE Computer Vision and Pattern Recognition (CVPR) conference — the largest international meeting devoted to artificial intelligence, machine learning, and computer vision research and applications.
Kavita Bala, dean of the Cornell Ann S. Bowers College of Computing and Information Science, gave a featured keynote speech to a packed house of thousands of attendees. Bharath Hariharan, assistant professor of computer science, received the prestigious 2022 Pattern Analysis and Machine Intelligence (PAMI) Young Researcher Award.
The CVPR conference is the premier meeting for computer vision and pattern recognition — a field that is growing exponentially, with more than 10,000 registrants this year. The meeting is highly selective, and organizers typically accept less than 30% of submitted papers and less than 5% of oral presentations. According to Google Scholar Metrics, CVPR is the highest impact computing venue.
On June 23, Bala gave a speech entitled “Understanding Visual Appearance from Micron to Global Scale.” In her presentation, she described her group's research on better visual understanding. This work includes graphics models for realistic visual appearance and rendering, reconstruction of shape and materials, and visual search and recognition for world-scale discovery of visual patterns and trends across geography and time.
Through this work, Bala said she aims to “build a deeper understanding of the appearance of individual objects, and also build our collective understanding of world-scale events as recorded through visual media.”
On June 21, the IEEE Computer Society announced Hariharan as the 2022 PAMI Young Researcher Award recipient. The honor is given annually in recognition of “a distinguished research contribution in computer vision” made by an early-career researcher within seven years of completing their Ph.D. Candidates are nominated by fellow researchers within the computer vision community,
"I am honored to be recognized by IEEE for my contributions to the field of computer vision and proud to continue Cornell's tradition of impactful research in this area," he said.
Hariharan’s work lies at the intersection of computer vision and machine learning. Currently, he is developing computer recognition systems capable of identifying specific objects and visual phenomena using a small number of training images and with very little or no supervision. Most traditional visual recognition systems require millions of carefully curated and annotated images to correctly identify visual concepts. Collecting these large training data sets can be challenging and expensive, which makes computer vision technology inaccessible for most applications. Ultimately, Hariharan seeks to make this technology available to all users by requiring minimal training data.
In March of this year, Hariharan also received the National Science Foundation (NSF) Faculty Early Career Development Award.
Patricia Waldron is a science writer for the Cornell Ann S. Bowers College of Computing and Information Science.