Feature-Space Transformation Improves Supervised Segmentation Across Scanners.
Annegreet van OpbroekHakim C. AchterbergMarleen de BruijnePublished in: MLMMI@ICML (2015)
Keyphrases
- feature space
- image segmentation
- segmentation algorithm
- segmentation method
- level set
- fully unsupervised
- feature selection
- high dimensional
- segmentation accuracy
- multiscale
- unsupervised learning
- semi supervised
- supervised learning
- principal component analysis
- input space
- dimensionality reduction
- region growing
- energy function
- fully automatic
- medical images
- data points
- image retrieval
- high dimensionality
- range data
- edge detection
- object segmentation
- kernel function
- feature vectors
- input data
- shape prior
- dissimilarity measure
- linear transformation
- high dimensional feature space
- image analysis