Developers of ADAS and driverless vehicle tech now have access to a bigger thermal data set from Teledyne FLIR. The company’s starter data set, which is said to be the first free data set in the automotive industry, has now doubled in size.
It features more than 26,000 annotated images from the USA, England and France during the day and night time. It has also tripled the amount of categories. Furthermore, labels now include person, bike, car, motorcycle, bus, train, truck, traffic light, fire hydrant, street sign, dog, skateboard, stroller scooter, and other vehicle labels. This will ultimately enable faster evaluation of vehicle safety algorithm performance, neural network testing, and analysis of thermal sensors, such as the FLIR AD.
“The industry-first free thermal data set has been used by developers and as a tool in academic research to explore safety of automated driving technology with thermal imaging. The expanded data set, with more images and categories, will further enhance detection and classifications models, especially focused on automatic braking capabilities as part of a multispectral, sensor-fused system,” commented Chris Posch, director of automotive engineering at Teledyne FLIR.
”When combined with visible light cameras, lidar and radar, thermal sensors help create a comprehensive and redundant system to identify and classify roadway objects using sensor fusion data. Testing has demonstrated that thermal sensors are uniquely capable of seeing pedestrians, large animals and other vulnerable road users in conditions where current AEB systems are challenged, including in total darkness, most fog, smoke, shadows, inclement weather, and sun or headlight glare.”