Loading...
Loading...

Go to the content (press return)

Visual semantic re-ranker for text spotting

Author
Sabir, A.; Moreno-Noguer, F.; Padro, L.
Type of activity
Presentation of work at congresses
Name of edition
23rd Iberoamerican Congress on Pattern Recognition
Date of publication
2018
Presentation's date
2019-03-03
Book of congress proceedings
Proceedings of 23rd Iberoamerican Congress on Pattern Recognition
First page
884
Last page
892
DOI
10.1007/978-3-030-13469-3_102
Project funding
HuMoUR: Markerless 3D human motion understanding for adaptive robot behavior - TIN2017-90086-R
Repository
http://hdl.handle.net/2117/132950 Open in new window
URL
https://link.springer.com/chapter/10.1007%2F978-3-030-13469-3_102 Open in new window
Abstract
Many current state-of-the-art methods for text recognition are based on purely local information and ignore the semantic corre- lation between text and its surrounding visual context. In this paper, we propose a post-processing approach to improve the accuracy of text spotting by using the semantic relation between the text and the scene. We initially rely on an off-the-shelf deep neural network that provides a series of text hypotheses for each input image. These text hypotheses are then re-ran...
Keywords
Computer vision, Deep learning, Semantic visual context, Text spotting
Group of research
GPLN - Natural Language Processing Group
IDEAI-UPC - Intelligent Data Science and Artificial Intelligence Research Center
ROBiri - IRI Robotics Group
TALP - Centre for Language and Speech Technologies and Applications

Participants