Abstract:
In this talk, I will show that these are possible and discuss our proposed mathematical pipeline and methodology that has enabled, for instance, the first demonstration of person identification through walls from candidate video footage, or the first demonstration of crowd counting through walls, using only WiFi signals. More specifically, I will start by discussing XModal-ID, a WiFi-video cross-modal gait-based person identification system that can take a WiFi signal measured when an unknown person walks in an unknown area and a video footage of a walking person in another area, to determine whether it is the same person in both cases or not. Here I will show how to translate the video content to the RF domain and further develop a processing pipeline to extract key features from both signals for the purpose of identification. This video to RF pipeline has implications beyond person identification and in the general area of activity recognition, as we shall see. More specifically, I will show how it can enable training activity-recognition RF sensing systems without any real training data. Next, I will discuss our proposed methodology for occupancy analytics with WiFi signals. Here, we shall see how parameters such as the PDF of the received signal power or the PDF of the inter-event times carry vital information on the crowd, which I will theoretically characterize. I will then show how crowd information such as crowd count, crowd speed, and arrival/departure rate, can be robustly extracted from only the received WiFi power measurements. It is noteworthy that our proposed approaches for person identification or crowd counting are not based on collecting training data of the people involved in the experiments. Finally, I will show that it is indeed possible to image still objects through walls and discuss how our methodology for the co-optimization of path planning and communication has enabled the first demonstration of 3D imaging through walls with only drones and WiFi.
Bio:
Yasamin Mostofi received a B.S. degree in electrical engineering from Sharif University of Technology, and M.S. and Ph.D. degrees from Stanford University. She is currently a professor in the Department of Electrical and Computer Engineering at the University of California Santa Barbara. Yasamin is the recipient of the 2016 Antonio Ruberti Prize from the IEEE Control Systems Society, the Presidential Early Career Award for Scientists and Engineers (PECASE), the National Science Foundation (NSF) CAREER award, and the IEEE 2012 Outstanding Engineer Award of Region 6 (more than 10 Western U.S. states), among other awards. She was a semi-plenary speaker at the 2018 IEEE Conference on Decision and Control (CDC) and a keynote speaker at the 2018 Mediterranean Conference on Control and Automation (MED). Her current research thrusts include RF sensing, X-ray vision for robots, occupancy analytics with everyday communication signals, see-through imaging and person identification with WiFi, communication-aware robotics, and human-robot networks. Her research has appeared in several reputable news venues such as BBC, Huffington Post, Daily Mail, Engadget, TechCrunch, NSF Science360, ACM News, and IEEE Spectrum, among others. She is a fellow of IEEE.
