HW-TSC at IWSLT2023: Break the Quality Ceiling of Offline Track via Pre-Training and Domain Adaptation.
Zongyao LiZhanglin WuZhiqiang RaoYuhao XieJiaxin GuoDaimeng WeiHengchao ShangMinghan WangXiaoyu ChenZhengzhe YuLi ShaoJunLei LiZhiHao YangPublished in: IWSLT@ACL (2023)
Keyphrases
- domain adaptation
- covariate shift
- cross domain
- semi supervised
- multiple sources
- training data
- transfer learning
- labeled data
- semi supervised learning
- unlabeled data
- natural language processing
- manifold alignment
- test set
- co training
- target domain
- active learning
- training process
- information retrieval
- natural language
- training set
- document classification
- learning tasks
- test data
- supervised learning
- training examples
- training samples