The Cornell Undergraduate Vision and Learning (CUVL) club had its first major conference paper accepted to the International Conference in Computer Vision (ICCV 2019). The paper—“Enhancing Adversarial Example Transferability with an Intermediate Level Attack”—was written by Cornell undergraduates Qian Huang, Isay Katsman, Zeqi Gu, and Horace He under the direction of CS Professor Serge Belongie and Ser-Nam Lim (of Facebook AI). The paper entered a dizzyingly vast field of submissions—some 4,303 in all.
The research conducted in this paper is related to CUVL’s collaboration with Facebook, and employed Graphics Processing Units (GPUs) purchased with funds from a grant received from Facebook last year.
Cornell University Vision and Learning (CUVL) focuses on undergraduate research and publication, as well as educating undergraduates on state-of-the-art Vision/Machine Learning research. The program is sponsored by and in direct collaboration with Facebook AI.
Belongie notes that Katsman and He will use this acceptance (along with a recent workshop paper acceptance) to expand the information at the current CUVL website. Please check back for further developments. And in the meantime, if you would like to receive updates from CUVL, please join the mailing list.
Belongie, Assistant Professor Bharath Hariharan, and Professor and Chair Kavita Bala, serve as affiliated faculty for CUVL. Professor Bala is very proud of the undergraduates in this program, noting their exceptional commitment to taking up leading-edge challenges. The confirmation of their good work by ICCV 2019 is another indication of their progress toward contributing to the future of computer vision and learning.