An automated system for semantic object labeling with soft object recognition and dynamic programming segmentation.
Jonas ClevelandDinesh ThakurPhilip M. DamesCody J. PhillipsTerry KientzKostas DaniilidisJohn BergstromVijay KumarPublished in: CASE (2015)
Keyphrases
- object recognition
- dynamic programming
- image segmentation
- object segmentation
- d objects
- multiple objects
- graph cut segmentation
- bounding box
- object detection
- object categories
- figure ground segmentation
- segmentation method
- segmentation algorithm
- man made objects
- object matching
- object representation
- object model
- object shape
- level set
- region growing
- image regions
- multiscale
- image segments
- computer vision
- final segmentation
- pixel level
- rib cage
- interactive image segmentation
- object contours
- semantic segmentation
- low level features
- fully automatic
- object description
- signed distance
- natural language
- stereo matching
- object class
- labeling process
- image labeling
- automated segmentation
- semantic information
- high level
- recognizing objects
- active learning
- global shape
- automated analysis
- object detectors
- scene understanding
- shape prior
- object tracking
- energy function
- semantic web
- natural images
- foreground and background
- target object
- object classes
- spatial relations
- image segmentations
- image classification
- medical images
- image features