Efficient Classification of Very Large Images with Tiny Objects.
Fanjie KongRicardo HenaoPublished in: CoRR (2021)
Keyphrases
- image classification
- image data
- complex scenes
- image pixels
- image regions
- real world scenes
- image database
- real world objects
- individual objects
- image registration
- target object
- bounding box
- image features
- multiple objects
- visual appearance
- real objects
- object recognition
- pattern recognition
- partial occlusion
- image collections
- spatial relationships
- d objects
- hyperspectral images
- complex background
- training set
- objects represented
- visual data
- three dimensional
- preprocessing stage
- keypoints
- edge detection
- input image
- ground truth
- image analysis
- multiple images
- pigmented skin lesions
- thematic mapper
- spatially localized
- decision trees
- feature extraction
- similarity measure
- support vector
- classification accuracy
- background clutter
- extracted features
- text classification
- geometric constraints
- lighting conditions
- image annotation
- three dimensional objects
- image retrieval
- skin lesion
- detecting objects
- image set
- object classes