Transfer-Once-For-All: AI Model Optimization for Edge
Achintya Kundu, Laura Wynter, et al.
EDGE 2023
We present a novel and effective method for determining the placement of sensors so as to be able to satisfy probabilistic constraints on the time-to-detection of an incident. Indeed, with the wealth of real-time traffic data available today, an important new goal of intelligent traffic management systems is incident detection with time-to-detection guarantees, in particular on expressways with large distances between sensors. This goal drives investment decisions in new sensor deployment, hence making the topic a pressing need for traffic management. The method we provide makes use of a probabilistic formulation of traffic behavior and incident localization to determine the minimum spacing of sensors needed to achieve the time-to-detection goal with a specified probability.
Achintya Kundu, Laura Wynter, et al.
EDGE 2023
Supriyo Ghosh, Laura Wynter, et al.
UAI 2022
Wanli Min, Laura Wynter
Transp. Res. Part C Emerg. Technol.
Fabian Lim, Laura Wynter, et al.
ICML 2022