DeepMVS: Learning Multi-view Stereopsis

被引：367

作者：

Huang, Po-Han ^{[1
]}

Matzen, Kevin ^{[2
]}

Kopf, Johannes ^{[2
]}

Ahuja, Narendra ^{[1
]}

Huang, Jia-Bin ^{[3
]}

机构：

[1] Univ Illinois, Urbana, IL 61801 USA

[2] Facebook, Menlo Pk, CA USA

[3] Virginia Tech, Blacksburg, VA USA

来源：

2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR) | 2018年

关键词：

D O I：

10.1109/CVPR.2018.00298

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We present DeepMVS, a deep convolutional neural network (ConvNet) for multi-view stereo reconstruction. Taking an arbitrary number of posed images as input, we first produce a set of plane-sweep volumes and use the proposed DeepMVS network to predict high-quality disparity maps. The key contributions that enable these results are (1) supervised pretraining on a photorealistic synthetic dataset, (2) an effective method for aggregating information across a set of unordered images, and (3) integrating multi-layer feature activations from the pre-trained VGG-19 network. We validate the efficacy of DeepMVS using the ETH3D Benchmark. Our results show that DeepMVS compares favorably against state-of-the-art conventional MVS algorithms and other ConvNet based methods, particularly for neartextureless regions and thin structures.

引用

页码：2821 / 2830

页数：10