An Approach Based on Transformer and Deformable Convolution for Realistic Handwriting Samples Generation.
Yiming WangHeng WangShiwen SunHongxi WeiPublished in: ICPR (2022)
Keyphrases
- fuzzy logic
- data sets
- image processing
- training set
- real life
- generation process
- handwritten characters
- fault diagnosis
- deformable models
- medical images
- incipient fault
- convolution kernel
- data samples
- articulated objects
- character recognition
- hidden markov models
- computer vision
- genetic algorithm
- real world
- neural network