This is an old revision of the document!


Bridging text spotting and SLAM with junction features

Wang, H.-C., Finn, C., Paull, L., Kaess, M., Rosenholtz, R., Teller, S., & Leonard, J.


Abstract

Navigating in a previously unknown environment and recognizing naturally occurring text in a scene are two important autonomous capabilities that are typically treated as distinct. However, these two tasks are potentially complementary, (i) scene and pose priors can benefit text spotting, and (ii) the ability to identify and associate text features can benefit navigation accuracy through loop closures. Previous approaches to autonomous text spotting typically require significant training data and are too slow for real-time implementation. In this work, we propose a novel high-level feature descriptor, the “junction”, which is particularly well-suited to text representation and is also fast to compute. We show that we are able to improve SLAM through text spotting on datasets collected with a Google Tango, illustrating how location priors enable improved loop closure with text features.

===== Information =====

title:
Bridging text spotting and SLAM with junction features
author:
H.-C. Wang,
C. Finn,
L. Paull,
M. Kaess,
R. Rosenholtz,
S. Teller,
& J. Leonard
citation:
IEEE/RSJ International Conference on Intelligent Robots and Systems
shortcite:
IROS
year:
2015
created:
2016-05-10
summary:
textspottingslam
tag:
rosenholtz
pdf:
http://people.csail.mit.edu/lpaull/publications/Wang_IROS_2015.pdf
type:
publication
 
publications/textspottingslam.1462912344.txt.gz · Last modified: 2016/05/10 16:32 by rosenholtz