Con-text: text detection using background connectivity for fine-grained object classification.
Sezer KaraogluJan C. van GemertTheo GeversPublished in: ACM Multimedia (2013)
Keyphrases
- fine grained
- text detection
- object classification
- object recognition
- text regions
- connected components
- text information
- complex background
- object detection
- object categories
- access control
- video analysis
- computer vision
- video frames
- face recognition
- bounding box
- document images
- image understanding
- text lines
- d objects
- pairwise
- text categorization
- image representation
- natural images
- input image
- three dimensional