Datasets

Stanford Multiview Light Field Dataset

This dataset was collected with a single hand-held Lytro Illum camera during my 2017 internship with the Stanford Computational Imaging. The dataset contains 4211 light fields organized into 30 categories (i.e. trees, benches), and each scene is captured from 4-6 camera poses.

Find the dataset here!

Download the research poster:

UNIZAR 360º Video Dataset with Depth Maps

This dataset was collected using a Kandao Obisidian camera during my 2019 internship with the Graphics and Imaging Lab at Universidad de Zaragoza in Spain. The dataset contains over 100 360º videos with depth maps generated for each one. This dataset includes videos with moving objects, such as cars and people, and static scenes, where no motion happens. Scenes vary in complexity, going from one moving person to three, and others have furniture that is moved farther and farther away from the camera to use in motion parallax testing.

The dataset and research poster will be published in the near future.