An image model based on occluding object images and maximum entropy.
John A. StullerRahul ShahPublished in: IEEE Trans. Image Process. (1998)
Keyphrases
- maximum entropy
- input image
- random fields
- image regions
- image data
- normalized correlation
- image collections
- image features
- test images
- region of interest
- lighting conditions
- image retrieval
- image classification
- multiple objects
- bounding box
- image set
- pixel level
- complex scenes
- target object
- keypoints
- segmentation method
- edge detection
- image segments
- image structure
- surface shape
- web images
- image content
- color histogram
- photometric properties
- maximum entropy principle
- segmented images
- feature points
- ground truth
- partial occlusion
- three dimensional
- image annotation
- foreground and background
- similarity measure
- conditional random fields
- class specific
- object segmentation
- image segmentation
- transformation based learning
- single image
- probabilistic model
- object recognition
- object shapes
- depth discontinuities
- minimum cross entropy
- iterative scaling
- d objects
- light source
- object classes