OmniDepth: Dense Depth Estimation for Indoors Spherical Panoramas

TL;DR

This paper introduces OmniDepth, a method for dense depth estimation in indoor spherical panoramas, leveraging a large synthetic dataset created from 3D data to train models specifically for 360 images.

Contribution

The work presents a novel approach to training depth estimation models directly on 360 datasets by re-purposing existing 3D datasets through rendering, enabling better performance on omnidirectional images.

Findings

Models trained on the synthetic dataset outperform those trained on traditional images.

The synthetic dataset is larger and more diverse than existing projective datasets.

Promising results are demonstrated on both synthesized and real-world 360 images.

Abstract

Recent work on depth estimation up to now has only focused on projective images ignoring 360 content which is now increasingly and more easily produced. We show that monocular depth estimation models trained on traditional images produce sub-optimal results on omnidirectional images, showcasing the need for training directly on 360 datasets, which however, are hard to acquire. In this work, we circumvent the challenges associated with acquiring high quality 360 datasets with ground truth depth annotations, by re-using recently released large scale 3D datasets and re-purposing them to 360 via rendering. This dataset, which is considerably larger than similar projective datasets, is publicly offered to the community to enable future research in this direction. We use this dataset to learn in an end-to-end fashion the task of depth estimation from 360 images. We show promising results in our synthesized data as well as in unseen realistic images.