Event registration banner
 

Deep Learning for Text Recognition

October 20, 2016
9:40 am - 10:00 am
GRB 330 A

Track: Artificial Intelligence
Type: Presentation
Level: Beginner/Intermediate

Text recognition using shallow features can be done with near human accuracy for images with constrained environments e.g. scanned documents. These features fail to achieve the same results when used outside such environments. In this work, we propose a hybrid approach to text recognition by leveraging information about the image type and sending different images through different optimized paths.

Speaker(s)

Manjari Akella, Machine Learning Engineer, IBM Watson