From Pixels to Random Walk Based Segments for Image Time Series Deep Classification.
Mohamed ChelaliCamille KurtzAnne PuissantNicole VincentPublished in: ICPRAI (2020)
Keyphrases
- random walk
- input image
- image pixels
- image classification
- pixel classification
- pixel values
- homogeneous regions
- multiscale
- image features
- directed graph
- image segmentation
- transition probabilities
- feature extraction
- image regions
- gibbs distribution
- image segments
- image retrieval
- machine learning
- markov chain
- spectral graph partitioning
- feature vectors
- flow graph
- transition probability matrix
- segmentation algorithm
- text classification
- foreground and background
- image collections
- graph laplacian
- pagerank algorithm