Nexar is a leader in the software development industry and a provider of the world’s first open vehicle-to-vehicle (V2V) network for preventing road collisions and enabling autonomous mobility. With over one billion insights collected each month, Nexar is paving the way for the utilization of AI in transportation data collection. In academic research, access to high-quality data is crucial to researchers testing theories and addressing critical challenges, which is why Nexar has announced the Nexar Dashcam Crash Prediction Challenge. This competition runs until May 4, 2025, with results being presented at CVPR 2025, the leading international conference on computer vision.
The Nexar Open Dataset is now live and offers researchers video samples of real-world driving scenarios, including collisions and near-collisions. This data is an exceptional resource for predictive modeling, innovations in road safety, and AI-driven mobility research. Your goal is to develop a model that accurately predicts vehicle collisions as early as possible in dashcam video sequences. The earlier and more accurately you can detect an impending accident, the better your score.
For more information on the competition, click the link below!