Deep Learning for Multimedia Processing Applications

CHAPTER 16 Challenges and Techniques to Improve Deep Detection and Recognition Methods for Text Spotting

Anuj Abraham^1,* and Shitala Prasad^2,*

¹Technology Innovation Institute, Abu Dhabi, United Arab Emirates

²Computer Science & Engineering, School of Mathematics & Computer Science, Indian Institute of Technology Goa, Goa, India

^* Both authors contributed equally to this work.

DOI: 10.1201/9781032646268-16

16.1 Introduction

Over the few years, the detection of text contents on images and presentation slides has attracted several researchers due to its potential applications, such as in video lecture and conference summarization, document analysis, robot assistance in stationery stores and shopping malls, forensic linguistics with text reading, ...