BFMC Computer Vision Project

This dataset was created to train YOLOv5 models. It focuses specifically on the field of 1:10 scale autonomous model vehicles as used in competitions such as the Bosch Future Mobility Challenge. It includes an extensive collection of images from a test track that emulates various urban driving conditions and shows a variety of objects such as vehicles, pedestrians, traffic signs and road markings that one might encounter in a scaled-down urban environment.

The images were captured in different lighting conditions to ensure that the trained models can perform reliably at different times of day. In addition, the dataset contains images from different viewpoints to cover a comprehensive range of object perspectives, which is crucial for the development of robust object recognition algorithms.

Each image in the dataset is annotated with bounding boxes and the respecting class labels for all objects depicted, providing accurate object recognition and classification. This dataset can be a useful resource for teams participating in scaled autonomous vehicle competitions. It provides labels for all of the objects present in the Bosch Future Mobility challenge and can be used for the development, testing and refinement of perception algorithms in the field of miniature autonomous vehicles.