Dynamic topology and relevance learning SOM-based algorithm for image clustering tasks.
Heitor R. MedeirosFelipe D. B. de OliveiraHansenclever de F. BassaniAluizio F. R. AraújoPublished in: Comput. Vis. Image Underst. (2019)
Keyphrases
- k means
- learning algorithm
- self organizing maps
- test images
- region of interest
- detection algorithm
- unsupervised manner
- clustering method
- estimation algorithm
- data clustering
- multiscale
- image processing algorithms
- image data
- kohonen self organizing map
- unsupervised learning
- segmentation algorithm
- topology preserving
- segmentation method
- single image
- affinity measure
- high resolution
- neural gas
- neural network
- energy function
- supervised learning
- reinforcement learning
- image regions
- image matching
- clustering algorithm
- image processing
- matching algorithm
- gray value
- hierarchical clustering
- kohonen self organizing maps
- spectral clustering
- cluster analysis
- image segmentation
- distance metric
- feature vectors
- optimal solution
- image representation
- expectation maximization
- edge detection
- active learning
- image retrieval