SurfaceNet: An End-to-end 3D Neural Network for Multiview Stereopsis

Abstract

This paper proposes an end-to-end learning framework for multiview stereopsis. We term the network SurfaceNet. It takes a set of images and their corresponding camera parameters as input and directly infers the 3D model. The key advantage of the framework is that both photo-consistency as well geometric relations of the surface structure can be directly learned for the purpose of multiview stereopsis in an end-to-end fashion. SurfaceNet is a fully 3D convolutional network which is achieved by encoding the camera parameters together with the images in a 3D voxel representation. We evaluate SurfaceNet on the large-scale DTU benchmark.

2017 iccv poster

Cited by in corpus (12)

• A Survey on Deep Learning Techniques for Stereo-based Depth Estimation
• Mini-Unmanned Aerial Vehicle-Based Remote Sensing: Techniques, Applications, and Prospects
• Robust Attentional Aggregation of Deep Feature Sets for Multi-view 3D Reconstruction
• Dense 3D Object Reconstruction from a Single Depth View
• 3D Sketching using Multi-View Deep Volumetric Prediction
• GIFS: Neural Implicit Function for General Shape Representation
• GC-MVSNet: Multi-View, Multi-Scale, Geometrically-Consistent Multi-View Stereo
• SurfaceNet+: An End-to-end 3D Neural Network for Very Sparse Multi-view Stereopsis
• Deep Multi-View Stereo gone wild
• Incremental Dense Reconstruction from Monocular Video with Guided Sparse Feature Volume Fusion
• Lightweight and Accurate Multi-View Stereo with Confidence-Aware Diffusion Model
• Cutting-edge 3D reconstruction solutions for underwater coral reef images: A review and comparison