I3CL: Intra- and Inter-Instance Collaborative Learning for Arbitrary-Shaped Scene Text Detection.
Bo DuJian YeJing ZhangJuhua LiuDacheng TaoPublished in: Int. J. Comput. Vis. (2022)
Keyphrases
- text detection
- arbitrary shaped
- collaborative learning
- urban scenes
- outdoor scenes
- natural scene images
- scene understanding
- d scene
- object recognition
- video analysis
- urban environments
- image sequences
- arbitrary shape
- complex scenes
- video sequences
- single image
- text information
- connected components
- range data
- video database
- vanishing points
- dynamic scenes
- point cloud
- multi band
- aerial images
- learning process
- camera calibration
- e learning
- moving objects
- planar surfaces
- photorealistic
- input image
- packing problem
- high accuracy
- multiple objects