Model Guidance via Explanations Turns Image Classifiers into Segmentation Models.
Xiaoyan YuJannik FranzenWojciech SamekMarina M.-C. HöhneDagmar KainmuellerPublished in: CoRR (2024)
Keyphrases
- probabilistic model
- image segmentation
- energy function
- multiscale
- bayesian framework
- parametric models
- classification models
- input image
- global shape
- image regions
- statistical model
- prior model
- prior information
- segmentation method
- image analysis
- autoregressive
- energy functional
- segmentation algorithm
- image dependent
- grey level
- similarity measure
- learned models
- parameter estimation
- image representation
- feature points
- markov random field
- image features
- expectation maximization
- object segmentation
- edge detection
- object models
- segmented images
- foreground and background
- image retrieval
- training set
- support vector
- training data
- explanatory power