Normalized Cut Loss for Weakly-Supervised CNN Segmentation.
Meng TangAbdelaziz DjelouahFederico PerazziYuri BoykovChristopher SchroersPublished in: CVPR (2018)
Keyphrases
- normalized cut
- weakly supervised
- image segmentation
- graph partitioning
- superpixels
- spectral clustering
- object segmentation
- segmentation algorithm
- clustering algorithm
- topic models
- object class
- mean shift
- graph cuts
- semi supervised
- segmentation method
- named entities
- weighted graph
- multiscale
- graph structure
- object detection
- image processing
- energy function
- machine learning
- shape prior
- bounding box
- active learning
- pairwise
- object detectors