Objectron
A large-scale dataset of object-centric video clips with 3D bounding box annotations and AR metadata for 3D object detection research.
What is Objectron?
Objectron is a large-scale dataset of short, object-centric video clips designed for 3D object detection and understanding research. It provides annotated videos with 3D bounding boxes and AR metadata like camera poses and point clouds, capturing everyday objects from multiple angles. The dataset addresses the need for high-quality 3D perception data to train and evaluate computer vision models.
Computer vision researchers, machine learning engineers, and developers working on 3D object detection, augmented reality, or robotic perception systems. It is particularly valuable for those needing multi-view object data with precise spatial annotations.
Developers choose Objectron for its scale, rich metadata, and ready-to-use formats that streamline 3D model training. Its unique combination of video sequences, 3D annotations, and AR session data provides a comprehensive foundation for advancing real-world 3D perception beyond 2D image datasets.
Overview
Objectron is a dataset of short, object-centric video clips. In addition, the videos also contain AR session metadata including camera poses, sparse point-clouds and planes. In each video, the camera moves around and above the object and captures it from different views. Each object is annotated with a 3D bounding box. The 3D bounding box describes the object’s position, orientation, and dimensions. The dataset contains about 15K annotated video clips and 4M annotated images in the following categories: bikes, books, bottles, cameras, cereal boxes, chairs, cups, laptops, and shoes
Use Cases
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
