Knowledge Guided Attention and Inference for Describing Images Containing Unseen Objects.
Aditya MogadalaUmanga BistaLexing XieAchim RettingerPublished in: ESWC (2018)
Keyphrases
- image data
- image regions
- real world objects
- ground truth
- complex scenes
- input image
- visual appearance
- geometric constraints
- image database
- three dimensional
- spatial relationships
- multiple objects
- image features
- image collections
- multiple images
- target object
- image analysis
- individual objects
- image pixels
- objects represented
- focus of attention
- real objects
- previously unseen
- image retrieval
- image registration
- object shape
- knowledge representation
- outdoor scenes
- rigid body
- visual data
- bounding box
- spatial information
- real world scenes
- knowledge base
- keypoints
- acquisition process
- lighting conditions
- geometric information
- image set
- image content
- d objects
- partial occlusion
- edge detection
- object detectors
- image segments
- object recognition
- moving objects
- viewing angle
- image processing
- domain knowledge
- background clutter
- natural images
- computer graphics
- image matching
- visual attention
- test images