Image Semantic Segmentation Using Deep Convolutional Nets, Fully Connected Conditional Random Fields, and Dilated Convolution.
Degui XiaoPei ZhongPublished in: HPCC/SmartCity/DSS (2019)
Keyphrases
- conditional random fields
- semantic segmentation
- fully connected
- random fields
- graphical models
- superpixels
- markov random field
- probabilistic model
- hidden markov models
- long range
- higher order
- information extraction
- pairwise
- generative model
- named entity recognition
- segmentation method
- information retrieval
- feature vectors
- image representation
- multiscale
- image segmentation
- similarity measure
- learning algorithm
- test images
- active contours
- image features