Visual and semantic ensemble for scene text recognition with gated dual mutual attention.
Zhiguang LiuLiangwei WangJian QiaoPublished in: Int. J. Multim. Inf. Retr. (2022)
Keyphrases
- scene text recognition
- selective attention
- high level
- object recognition
- visual information
- semantic information
- visual concepts
- neural network
- ensemble learning
- semantic similarity
- low level
- training data
- natural language
- semantic space
- visual features
- visual attention
- ensemble methods
- semantic concepts
- semantic content
- semantic knowledge
- data sets
- linear programming
- low level features
- image classification
- semantic analysis
- visual perception