Weakly Supervised Spatial Deep Learning for Earth Image Segmentation Based on Imperfect Polyline Labels.
Zhe JiangWenchong HeMarcus Stephen KirbyArpan Man SainjuShaowen WangLawrence V. StanislawskiEthan ShaversE. Lynn UseryPublished in: ACM Trans. Intell. Syst. Technol. (2022)
Keyphrases
- deep learning
- weakly supervised
- image segmentation
- superpixels
- topic models
- relation extraction
- object class
- spatial information
- graph cuts
- markov random field
- semi supervised
- object segmentation
- segmentation algorithm
- named entities
- multiscale
- automatic extraction
- deformable models
- image processing
- pairwise
- segmentation method
- object recognition
- unsupervised learning
- natural images
- level set
- domain specific
- training set
- computer vision