Learning Single-Image Depth from Videos using Quality Assessment Networks
arXiv:1806.09573
Abstract
Depth estimation from a single image in the wild remains a challenging problem. One main obstacle is the lack of high-quality training data for images in the wild. In this paper we propose a method to automatically generate such data through Structure-from-Motion (SfM) on Internet videos. The core of this method is a Quality Assessment Network that identifies high-quality reconstructions obtained from SfM. Using this method, we collect single-view depth training data from a large number of YouTube videos and construct a new dataset called YouTube3D. Experiments show that YouTube3D is useful in training depth estimation networks and advances the state of the art of single-view depth estimation in the wild.
References in corpus (15)
- PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
- Learning a Probabilistic Latent Space of Object Shapes via 3D Generative-Adversarial Modeling
- DeMoN: Depth and Motion Network for Learning Monocular Stereo
- ScanNet: Richly-annotated 3D Reconstructions of Indoor Scenes
- SfM-Net: Learning of Structure and Motion from Video
- Matterport3D: Learning from RGB-D Data in Indoor Environments
- Unsupervised CNN for Single View Depth Estimation: Geometry to the Rescue
- Learning a Multi-View Stereo Machine
- Deep Ordinal Regression Network for Monocular Depth Estimation
- SceneNet RGB-D: 5M Photorealistic Images of Synthetic Indoor Trajectories with Ground Truth
- Semi-Supervised Deep Learning for Monocular Depth Map Prediction
- Semantic Scene Completion from a Single Depth Image
- Unsupervised Monocular Depth Estimation with Left-Right Consistency
- Monocular Dense 3D Reconstruction of a Complex Dynamic Scene from Two Perspective Frames
- Self-Supervised Relative Depth Learning for Urban Scene Understanding