Objectron: A Large Scale Dataset of Object-Centric Videos in the Wild with Pose Annotations

TL;DR

Objectron is a large-scale dataset of object-centric videos with pose annotations designed to advance 3D object detection and related tasks, enabling new research in robotics, AR, and image retrieval.

Contribution

The paper introduces the Objectron dataset with 4 million images and pose annotations across nine categories, along with a new 3D detection evaluation metric.

Findings

Baseline models trained on Objectron show improved 3D detection performance.

The dataset facilitates research in 3D object tracking and view synthesis.

The new metric, 3D Intersection over Union, provides a standardized evaluation method.

Abstract

3D object detection has recently become popular due to many applications in robotics, augmented reality, autonomy, and image retrieval. We introduce the Objectron dataset to advance the state of the art in 3D object detection and foster new research and applications, such as 3D object tracking, view synthesis, and improved 3D shape representation. The dataset contains object-centric short videos with pose annotations for nine categories and includes 4 million annotated images in 14,819 annotated videos. We also propose a new evaluation metric, 3D Intersection over Union, for 3D object detection. We demonstrate the usefulness of our dataset in 3D object detection tasks by providing baseline models trained on this dataset. Our dataset and evaluation source code are available online at http://www.objectron.dev