Combining Bottom-Up, Top-Down, and Smoothness Cues for Weakly Supervised Image Segmentation.

Anirban Roy Sinisa Todorovic

Published in: CVPR (2017)

Keyphrases

bayesian inference
cost function
database
data sets
helmholtz principle
combining multiple
website
data mining
reinforcement learning
wide range
search algorithm
multiresolution
hidden markov models
probabilistic model
image processing
machine learning
smoothing parameter