Deep-learning-based object-level contour detection with CCG and CRF optimization.

Yuan Liao Songping Fu Xiaoqing Lu Chengcui Zhang Zhi Tang

Published in: ICME (2017)

Keyphrases

pixel level
conditional random fields
pairwise
superpixels
active learning
object recognition
training set
viewpoint
image features
multi class
object detection
graphical models
image classification
higher level
region segmentation