An innovative driver behavior prediction model was created using precise telemetric DataFromSky data by the researchers at the UC Berkeley This model was published in the paper ParkPredict+: Multimodal Intent and Motion Prediction for Vehicles in Parking Lots with CNN and Transformer paper by Xu Shen et al. You can read the full paper here.
What is more, along with this research a very unique FREE dataset was created that monitored 400 parking spots for 3.5 hours and provides full trajectory and bounding box data of all vehicles and even pedestrians. You can get the dataset here.
Looking for extremely precise telemetry data for your traffic project? Try our fully automated video analytics service TraffiSurvey for free and contact us!