Holopix50k: A Large-Scale In-the-wild Stereo Image Dataset

TL;DR

Holopix50k introduces a large-scale, in-the-wild stereo image dataset from mobile users, significantly enhancing stereo vision tasks and generalization in mobile photography applications.

Contribution

This paper presents Holopix50k, a new extensive stereo dataset collected from real-world mobile photography, improving training and performance of stereo vision algorithms.

Findings

Using Holopix50k improves stereo super-resolution results.

The dataset enhances self-supervised monocular depth estimation.

Holopix50k outperforms existing datasets in diversity and size.

Abstract

With the mass-market adoption of dual-camera mobile phones, leveraging stereo information in computer vision has become increasingly important. Current state-of-the-art methods utilize learning-based algorithms, where the amount and quality of training samples heavily influence results. Existing stereo image datasets are limited either in size or subject variety. Hence, algorithms trained on such datasets do not generalize well to scenarios encountered in mobile photography. We present Holopix50k, a novel in-the-wild stereo image dataset, comprising 49,368 image pairs contributed by users of the Holopix mobile social platform. In this work, we describe our data collection process and statistically compare our dataset to other popular stereo datasets. We experimentally show that using our dataset significantly improves results for tasks such as stereo super-resolution and self-supervised monocular depth estimation. Finally, we showcase practical applications of our dataset to motivate novel works and use cases. The Holopix50k dataset is available at http://github.com/leiainc/holopix50k