Congratulations to our Ph.D. student Abhijit Singh supervised by Associate Professor Biplab Sikdar, who has won the “Best Student Paper Award” at the IEEE 7th World Forum on the Internet of Things (WF-IoT) held from 14 June to 31 July 2021.
It included a broad program of papers and presentations on the latest technology developments and innovations in the many fields and disciplines that drive the utility and vitality of IoT solutions and applications. WF-IoT had strong involvement from the public sector and industry aimed at deepening the understanding, the necessary dialog, and actions needed to accelerate the adoption and deployment of IoT in various domains.
The winning paper was titled “Adversarial Attack for Deep Learning Based IoT Appliance Classification Techniques”. It showed that IoT applications which use machine learning may be susceptible to attacks by adversarial examples. He developed a white-box adversarial attack mechanism to generate adversarial examples for data obtained from smart meters installed in residential houses and demonstrated that their statistical properties were indistinguishable from those of the true data points. The attack mechanism focused specifically on deep learning-based models used to perform appliance classification in smart home environments. The statistical indistinguishability of the adversarial data points from the true data points indicates that non-ML based solutions may not be able to tackle the challenge posed by adversarial examples, and thus lightweight ML models which are deployable on edge devices need to be robust to such adversarial attacks, for safe large-scale adoption.
Abhijit’s research focuses on adversarial machine learning and currently working on black-box adversarial attacks and defenses for deep learning-based models using smart-meter data. The motivation of the research is to improve the understanding of these models and make them suitably robust for large-scale use. In the future, he also hopes to address privacy concerns arising due to the use of smart-meter data in training commercial machine learning models.
Details about the award – this paper was placed 2nd in the Best Student Paper Award. Abhijit received a letter of acknowledgment, a plaque, and a monetary award of 300 US dollars.
