Na and Ko are Ph.D. students in the Georgia Tech School of Electrical and Computer Engineering (ECE) and are members of the Gigascale Reliable Energy-Efficient Nanosystem (GREEN) Lab. They are advised by ECE Professor Saibal Mukhopadhyay.

Na’s and Ko’s award-winning demo was entitled “Deep Security: Toward Robust Deep Learning.” A successful deep learning-based computer vision task is perceived as a key enabler for autonomous vehicles. However, there have been numerous reports that deep-learning classifiers are vulnerable to small input perturbations that have been carefully generated by adversaries. Vulnerabilities in deep learning can become potential threats to successful autonomous driving. 

The objective of this research is to build robust deep-learning classifiers for various adversarial attacks in order to better protect self-driving cars. To address this challenge, Na and Ko propose embedding space for both classification and low-level (pixel-level) similarity learning that will ignore unknown pixel level perturbation. 

They also propose cascade adversarial training, which transfers the knowledge of the end results of adversarial training. This proposed approach shows improved accuracy compared to the current state-of-the-art adversarial training and ensemble adversarial training methodologies.

Photo cutline: ECE Ph.D. students Taesik Na and Jong Hwan Ko are pictured in the center of this photo. To the left of Na are Judge Kevin Skapinetz of IBM Security and Judge Blake Patton of Tech Square Ventures. To the right of Ko are Judge and Founding Donor of CREATE-X Chris Klaus and IISP Director Wenke Lee.