Learning to Recover 3D Scene Shape from a Single Image

Abstract

Despite significant progress in monocular depth estimation in the wild, recent state-of-the-art methods cannot be used to recover accurate 3D scene shape due to an unknown depth shift induced by shift-invariant reconstruction losses used in mixed-data depth prediction training, and possible unknown camera focal length. We investigate this problem in detail, and propose a two-stage framework that first predicts depth up to an unknown scale and shift from a single monocular image, and then use 3D point cloud encoders to predict the missing depth shift and focal length that allow us to recover a realistic 3D scene shape. In addition, we propose an image-level normalized regression loss and a normal-based geometry loss to enhance depth prediction models trained on mixed datasets. We test our depth model on nine unseen datasets and achieve state-of-the-art performance on zero-shot dataset generalization. Code is available at: https://git.io/Depth

References in corpus (6)

• Depth Map Prediction from a Single Image using a Multi-Scale Deep Network
• Learning Depth from Single Monocular Images Using Deep Convolutional Neural Fields
• Joint 2D-3D-Semantic Data for Indoor Scene Understanding
• DIODE: A Dense Indoor and Outdoor DEpth Dataset
• DiverseDepth: Affine-invariant Depth Prediction Using Diverse Data
• Holopix50k: A Large-Scale In-the-wild Stereo Image Dataset