Recurrent Connections Aid Occluded Object Recognition by Discounting Occluders.

Markus Roland Ernst Jochen Triesch Thomas Burwick

Published in: CoRR (2019)

Keyphrases

object recognition
computer vision
object detection
object categories
partially occluded
image features

object classification
template matching
object class
change point detection
image understanding
d objects

recurrent neural networks
image representation
scene understanding
visual recognition
man made
unsupervised learning

biologically motivated
multi unit combinatorial auctions
image matching
image interpretation
object categorization
recognition scheme