Improving the Detection of Relations Between Objects in an Image Using Textual Semantics.
Dennis MedvedFangyuan JiangPeter ExnerMagnus OskarssonPierre NuguesKalle ÅströmPublished in: ICPRAM (Selected Papers) (2014)
Keyphrases
- bounding box
- cluttered scenes
- spatial relations
- image classification
- image pixels
- image features
- multiscale
- multiple objects
- input image
- foreground background separation
- image analysis
- image data
- image retrieval
- complex scenes
- image content
- single image
- road signs
- partial occlusion
- spatial relationships
- keypoints
- complex background
- target object
- image regions
- image segmentation
- image representation
- object detectors
- scene understanding
- hough transform
- cluttered background
- object detection
- individual objects
- relational structures
- segmentation method
- similar objects
- visual appearance
- temporal relations
- lighting conditions
- spatial information
- d objects
- man made
- object models
- intensity variations
- web images
- pixel values
- semantic meaning
- image collections
- detecting objects