Domain Adaptation via Mutual Information Maximization for Handwriting Recognition.
Pei TangLiangrui PengRuijie YanHaodong ShiGang YaoChangsong LiuJie LiYuqi ZhangPublished in: ICASSP (2022)
Keyphrases
- domain adaptation
- handwriting recognition
- mutual information maximization
- mutual information
- speech recognition
- semi supervised
- character recognition
- multiple sources
- cross domain
- sentiment classification
- word segmentation
- word recognition
- test data
- labeled data
- semi supervised learning
- transfer learning
- document classification
- target domain
- optical character recognition
- feature selection
- unlabeled data
- computer vision
- pattern recognition