Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-Scale Convolutional Architecture

被引:1918
作者
Eigen, David [1 ]
Fergus, Rob [1 ,2 ]
机构
[1] NYU, Courant Inst, Dept Comp Sci, New York, NY 10003 USA
[2] Facebook AI Res, Menlo Pk, CA USA
来源
2015 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV) | 2015年
关键词
D O I
10.1109/ICCV.2015.304
中图分类号
TP18 [人工智能理论];
学科分类号
140502 [人工智能];
摘要
In this paper we address three different computer vision tasks using a single multiscale convolutional network architecture: depth prediction, surface normal estimation, and semantic labeling. The network that we develop is able to adapt naturally to each task using only small modifications, regressing from the input image to the output map directly. Our method progressively refines predictions using a sequence of scales, and captures many image details without any superpixels or low-level segmentation. We achieve state-of-the-art performance on benchmarks for all three tasks.
引用
收藏
页码:2650 / 2658
页数:9
相关论文
共 38 条
[1]
[Anonymous], DISCRIMINATIVELY TRA
[2]
[Anonymous], 2014, ECCV
[3]
[Anonymous], NIPS WORKSH DEEP LEA
[4]
[Anonymous], 2014, ARXIV14116387
[5]
[Anonymous], 2014, CVPR
[6]
[Anonymous], TPAMI
[7]
[Anonymous], 2015, ARXIV150104537
[8]
[Anonymous], 2015, ICLR
[9]
[Anonymous], 2013, CVPR
[10]
[Anonymous], 2013, CVPR