This website hosts an up-to-date archive of publicly available datasets for autonomous driving research. To provide practical guidance on when to use which dataset, we categorize the datasets along a variety of dimensions: the type of autonomous driving task, the diversity and modality of data, type of annotations, etc. We also indicate whether training/testing data and benchmarks are provided. Further, we highlight those datasets that capture low-driveability scenes, i.e., scenes with more complexity, dynamics, and potential hazards.
This driveability dataset index is part of a larger study on the issue of driveability and the factors and metrics that impact this issue. The results of that study can be found in the following paper: Is it Safe to Drive? An Overview of Factors, Challenges, and Datasets for Driveability Assessment in Autonomous Driving. We hope that this study will encourage both targeted dataset collection and the proposal of novel driveability metrics that enhance the robustness of autonomous cars in adverse environments.
Citation details: If you find this resource useful in your research, please cite the associated paper: Junyao Guo, Unmesh Kurup and Mohak Shah. Is it Safe to Drive? An Overview of Factors, Challenges, and Datasets for Driveability Assessment in Autonomous Driving.
Contributing: If there are additional data sources that you feel should be added to this resource, or if you are releasing such a dataset and would like it to be added, please reach out to us at: driveability.index@gmail.com
Note: Currently, we only include datasets that are collected by on-board sensors of a vehicle running on public roads, contain at least camera or LiDAR data, and allow free open access. Welcome and feel free to explore!