FDTA: Fully Convolutional Scene Text Detection With Text Attention.
Yongcun CaoShuaisen MaHaichuan PanPublished in: IEEE Access (2020)
Keyphrases
- text detection
- urban scenes
- scene text
- outdoor scenes
- natural scene images
- text information
- object recognition
- text regions
- d scene
- video analysis
- scene images
- connected components
- scene understanding
- visual attention
- video sequences
- input image
- structure from motion
- urban environments
- natural scenes
- aerial images
- complex background
- planar surfaces
- single image
- dynamic scenes
- moving objects
- range data
- image sequences
- web pages
- photorealistic
- text lines
- natural images