CodeSLAM - Learning a Compact, Optimisable Representation for Dense Visual SLAM

File Description SizeFormat 
main.pdfFile embargoed until 01 January 100004.92 MBAdobe PDF    Request a copy
Title: CodeSLAM - Learning a Compact, Optimisable Representation for Dense Visual SLAM
Authors: Bloesch, M
Czarnowski, J
Clark, R
Leutenegger, S
Davison, AJ
Item Type: Conference Paper
Abstract: The representation of geometry in real-time 3D per- ception systems continues to be a critical research issue. Dense maps capture complete surface shape and can be augmented with semantic labels, but their high dimension- ality makes them computationally costly to store and pro- cess, and unsuitable for rigorous probabilistic inference. Sparse feature-based representations avoid these problems, but capture only partial scene information and are mainly useful for localisation only. We present a new compact but dense representation of scene geometry which is conditioned on the intensity data from a single image and generated from a code consisting of a small number of parameters. We are inspired by work both on learned depth from images, and auto-encoders. Our approach is suitable for use in a keyframe-based monocular dense SLAM system: While each keyframe with a code can produce a depth map, the code can be optimised efficiently jointly with pose variables and together with the codes of overlapping keyframes to attain global consistency. Condi- tioning the depth map on the image allows the code to only represent aspects of the local geometry which cannot di- rectly be predicted from the image. We explain how to learn our code representation, and demonstrate its advantageous properties in monocular SLAM.
Issue Date: 18-Jun-2018
Date of Acceptance: 19-Feb-2018
URI: http://hdl.handle.net/10044/1/58316
Publisher: IEEE
Copyright Statement: This paper is embargoed until publication.
Sponsor/Funder: Dyson Technology Limited
Funder's Grant Number: PO 4500378543
Conference Name: IEEE Computer Vision and Pattern Recognition 2018
Publication Status: Accepted
Start Date: 2018-06-18
Finish Date: 2018-06-22
Conference Place: Salt Lake City, Utah, USA
Embargo Date: publication subject to indefinite embargo
Appears in Collections:Faculty of Engineering
Computing



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons