gvnn: Neural Network Library for Geometric Computer Vision

被引：48

作者：

Handa, Ankur ^{[1
]}

Bloesch, Michael ^{[3
]}

Patraucean, Viorica ^{[2
]}

Stent, Simon ^{[2
]}

McCormac, John ^{[1
]}

Davison, Andrew ^{[1
]}

机构：

[1] Imperial Coll London, Dept Comp, Dyson Robot Lab, London, England

[2] Univ Cambridge, Dept Engn, Cambridge, England

[3] ETH, Robot Syst Lab, Zurich, Switzerland

来源：

COMPUTER VISION - ECCV 2016 WORKSHOPS, PT III | 2016年 / 9915卷

关键词：

Spatial transformer networks; Geometric vision; Unsupervised learning;

D O I：

10.1007/978-3-319-49409-8_9

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We introduce gvnn, a neural network library in Torch aimed towards bridging the gap between classic geometric computer vision and deep learning. Inspired by the recent success of Spatial Transformer Networks, we propose several new layers which are often used as parametric transformations on the data in geometric computer vision. These layers can be inserted within a neural network much in the spirit of the original spatial transformers and allow backpropagation to enable end-to-end learning of a network involving any domain knowledge in geometric computer vision. This opens up applications in learning invariance to 3D geometric transformation for place recognition, end-to-end visual odometry, depth estimation and unsupervised learning through warping with a parametric transformation for image reconstruction error.

引用

页码：67 / 82

页数：16