Tuesday, June 19, 2018 — FLIR today announced the upcoming release of a first-of-its-kind, free machine learning thermal dataset for Advanced Driver Assistance Systems (ADAS) and self-driving vehicle researchers, developers, and auto manufacturers. The starter dataset features a compilation of more than 10,000 annotated thermal images of people, cars, other vehicles, bicycles and dogs in day and nighttime scenarios, enabling developers to begin testing and evolving convolutional neural networks (CNN) while using the FLIR Automotive Development Kit (ADK™).
The starter thermal imaging dataset allows the automotive community to quickly evaluate thermal sensors on next-generation algorithms being created for self-driving vehicles. When combined with visible light cameras, LIDAR, and radar, thermal sensor data paired with machine learning, helps create a more comprehensive and redundant system for identifying and classifying roadway objects, especially pedestrians and other living things.
With over a decade of experience in the automotive industry, FLIR has more than 500,000 automotive-qualified FLIR thermal sensors in driver warning systems from automakers such as General Motors, Volkswagen, Audi, BMW, and Mercedes-Benz. FLIR thermal cameras have proven reliable in the classification of pedestrians, bicycles, and vehicles in challenging lighting conditions including total darkness, fog, smoke, shadows, inclement weather, and sun glare at nearly four times the distance of typical headlights.
Recent high-profile autonomous-driving related accidents demonstrate the need for affordable, intelligent thermal sensors. With the potential for millions of autonomous-enabled vehicles, FLIR anticipates the costs of its thermal sensor to decrease significantly, which will encourage wide-scale adoption and ultimately enable safer autonomous vehicles.