Deep Structured Output Learning for Unconstrained Text Recognition

We develop a representation suitable for the unconstrained recognition of words in natural images: the general case of no fixed lexicon and unknown length. To this end we propose a convolutional neural network (CNN) based architecture which incorporates a Conditional Random Field (CRF) graphical...

Full description

Bibliographic Details
Main Authors: Jaderberg, Max, Simonyan, Karen, Vedaldi, Andrea, Zisserman, Andrew
Format: text
Published: 2014
Subjects:
Online Access:http://arxiv.org/abs/1412.5903