Global and Local Features Through Gaussian Mixture Models on Image Semantic Segmentation.
Darwin SaireAdín Ramírez RiveraPublished in: IEEE Access (2022)
Keyphrases
- semantic segmentation
- gaussian mixture model
- image features
- low level
- mixture model
- scene classification
- input image
- single image
- feature vectors
- test images
- conditional random fields
- image segmentation
- image content
- object categories
- image set
- object classes
- spatial information
- keypoints
- color histogram
- image classification
- weakly supervised
- multiscale
- feature space
- superpixels
- segmentation method
- image retrieval
- similarity measure
- feature set
- image processing
- pixel level
- image matching
- generative model
- image representation
- graphical models
- markov random field
- maximum likelihood
- text mining
- training data
- hidden markov models
- feature selection
- prior knowledge
- image regions