Houghencoder: Neural Network Architecture for Document Image Semantic Segmentation.

Alexander Sheshku Dmitry Nikolaev Vladimir L. Arlazaro

Published in: ICIP (2020)

Keyphrases

document images
semantic segmentation
conditional random fields
superpixels
scene classification
weakly supervised
object categories
object classes
object recognition
image understanding
object class
image set
markov random field
image segmentation
higher order
natural scenes