Autonomous Vehicles_VSIM Simulator_Dataset

The VSIM dataset is a specially designed collection of 5,000 images, created to enhance object detection in autonomous vehicle (AV) applications. Developed with the VSIM simulator in Unity, this dataset captures eight distinct categories across realistic driving scenarios, with a focus on representing varied environments and conditions for AVs. To ensure high-quality data, images were preprocessed and augmented using Roboflow. The dataset is divided into training, validation, and testing sets, with a unique setup for federated learning, where training data is split across three clients. Each client is configured to detect four key object classes—humans, cars, road signs, and bikes—supporting research in both centralized and federated contexts. The VSIM dataset fills a gap in AV data diversity, providing a valuable resource for advancing real-time object detection in distributed learning applications.