MegaDepth: Learning Single-View Depth Prediction from Internet Photos
This addresses the data scarcity issue in computer vision for depth prediction, enabling more robust models across varied environments, though it is incremental by enhancing existing data generation methods.
The authors tackled the problem of limited training data for single-view depth prediction by creating MegaDepth, a large dataset derived from multi-view Internet photos using structure-from-motion and multi-view stereo, which improved model generalization to diverse datasets like Make3D, KITTI, and DIW without training on them.
Single-view depth prediction is a fundamental problem in computer vision. Recently, deep learning methods have led to significant progress, but such methods are limited by the available training data. Current datasets based on 3D sensors have key limitations, including indoor-only images (NYU), small numbers of training examples (Make3D), and sparse sampling (KITTI). We propose to use multi-view Internet photo collections, a virtually unlimited data source, to generate training data via modern structure-from-motion and multi-view stereo (MVS) methods, and present a large depth dataset called MegaDepth based on this idea. Data derived from MVS comes with its own challenges, including noise and unreconstructable objects. We address these challenges with new data cleaning methods, as well as automatically augmenting our data with ordinal depth relations generated using semantic segmentation. We validate the use of large amounts of Internet data by showing that models trained on MegaDepth exhibit strong generalization-not only to novel scenes, but also to other diverse datasets including Make3D, KITTI, and DIW, even when no images from those datasets are seen during training.