A curated collection of robotics and computer vision datasets for research and development.
Awesome Robotics Datasets is a curated GitHub repository listing publicly available datasets for robotics and computer vision research. It organizes datasets by application domain (like driving, flying, indoor) and research topic (like SLAM, object tracking, 3D reconstruction) to help researchers and engineers quickly find relevant data for their projects.
Robotics researchers, computer vision engineers, PhD students, and developers working on perception, navigation, or autonomous systems who need benchmark or training data.
It saves significant time by aggregating and categorizing hundreds of datasets from academic labs, companies, and competitions into a single, community-vetted resource with quality indicators.
A collection of useful datasets for robotics and computer vision
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Datasets are organized by environment (e.g., driving, flying) and research topic (e.g., SLAM, object tracking), as shown in the README's clear sections, making targeted browsing efficient.
Includes thumbs-up recommendations and notes on dataset status (e.g., broken links like Radish), helping users identify reliable and maintained sources based on community input.
Encompasses classic benchmarks to modern large-scale datasets from academia and industry, such as KITTI and Waymo Open Dataset, providing a wide range for various research needs.
Provides direct links to dataset pages maintained by leading research groups like TUM CVG and Oxford VGG, ensuring access to original and authoritative data sources.
The repository is solely a list of links; users must independently handle downloading, formatting, and preprocessing data from each source, which adds significant overhead.
Relies on community maintenance, and some links are already noted as not working (e.g., Radish), risking outdated or inaccessible resources without automatic updates.
With hundreds of datasets, there's no built-in search or filter beyond the categorical structure, making it cumbersome to find specific datasets without manual scanning.