Talk: RF-Visual Perception with Applications to Mobile Sensing, Robotics, & Augmented Reality (ONLINE)

LIVE STREAM: https://uwmadison.zoom.us/j/93475184247?pwd=NHAvYWs4ZEpJYXF3MGg4K1VLeVN1QT09

Abstract: This talk presents RF-Visual perception, a new approach for sensing, perceiving, and interacting with the physical world. Unlike classical perception approaches like computer vision, which struggle in non-line-of-sight or cluttered environments, RF-Visual perception works across line-of-sight and non-line-of-sight environments, through occlusions, and in low-lighting conditions. 

The talk will describe the fundamentals of RF-Visual perception and how I used it to design, build, and evaluate new robotics and augmented reality systems. The talk will also introduce multiple novel primitives in RF-Visual perception for various computational tasks including localization, servoing/navigation, robotic grasping, reinforcement learning, and human-computer interaction. The talk will conclude by describing how this perception technology has already been deployed in practice and how the fundamental primitives described pave the way for future full-spectrum perception with applications in mobile sensing, robotics, HCI, environmental monitoring, and more. 

Bio: Tara Boroushaki is a Ph.D student at MIT, advised by Fadel Adib. Her research interests lie in sensing and mobile technologies with applications in wireless networking, cyber-physical systems, cyber-human systems, and human-computer interaction (HCI). Boroushaki’s research has received multiple awards including the IEEE RFID Best Paper Award, ACM SenSys Best Paper Honorable Mention, and the Microsoft Research PhD Fellowship. Her work has also been chosen as one of the top 10 AI Trends to Watch by The Wall Street Journal and one of the “100 Ways MIT is making the World Better” by MIT. Her research has had commercial impact through a startup, Cartesian Systems, that is deploying her technologies to address problems in retail and supply chain.