Superpixel clustering with deep features for unsupervised road segmentation.
Shunta SaitoTommi KerolaSatoshi TsutsuiPublished in: CoRR (2017)
Keyphrases
- unsupervised feature selection
- unsupervised learning
- feature extraction
- image segmentation
- clustering algorithm
- agglomerative clustering
- segmentation accuracy
- similarity function
- data clustering
- medical images
- k means
- unsupervised clustering
- multiscale
- bayesian information criterion
- unsupervised classification
- global shape
- image segmentation algorithm
- low level
- feature set
- segmentation method
- gabor filters
- normalized cut
- region growing
- deep learning
- fully unsupervised
- information theoretic
- text segmentation
- image regions
- object recognition