Cost-effective and accurate solutions are needed to detect pedestrians during the day and especially at night to implement safety measures. This year’s challenge is In partnership with the City of San José’s Vision Zero program. Vision Zero is the city’s initiative to significantly reduce traffic fatalities and injuries. People killed while walking are the biggest group of traffic fatalities and a significant number of fatalities occur at night. In fact, traffic fatalities have more than doubled since 2012 in San José. Tiny ML is a new technology that allows Machine Learning (ML) models to run on low-power microcontrollers that do not require a network infrastructure. We believe that TinyML has a great role to play in building systems that improve our world for everyone.
Proud to lead team NeurOHM from Technische Hochschule Nürnberg Georg Simon Ohm towards winning second place in the Pedestrian Detection Competition of tinyML Foundation. Our solution, Spiking Perception and processing for Intelligent Detection of pEdestrians on urban Roads (SPIDER) accurately detected and tracked pedestrians and bikers, while preserving their privacy, on the streets of Nuremberg day and night only using 5W. Neuromorphic sensing and processing made by SPICES Lab.