When you label objects and their location in an image, the tool uses the labels to build a 3D model of the scene. The tool does not require from you any knowledge about geometry, as all of the 3D information is automatically inferred from the annotations. For instance, the tool will know that a 'road' is a horizontal surface and that a 'car' is supported by the road. The tool learns to go from 2D to 3D using all the other labels already present in the database. The more images that are labeled, the better models the tool will learn.

By using this tool, you can help us build a large collection of annotated images for computer vision research. Currently, computers have a difficult time recognizing objects in images. While practical solutions exist for a few simple classes, such as human faces or cars, the more general problem of recognizing all different classes of objects in the world (e.g. guitars, bottles, telephones) remains unsolved. Computer vision researchers are currently investigating methods that can recognize and localize thousands of different object categories in complex scenes. A key component of these algorithms is the data used to train the computer's model of each object.

This work is the result of a collaboration between the Computer Science and Artificial Intelligence Laboratory at MIT and the INRIA (Willow project-team, Laboratoire d'Informatique de l'École Normale Supérieure, ENS/INRIA/CNRS UMR 8548). Funding for this research was provided by National Science Foundation Career award (IIS 0747120).

Label the objects