Multitask Learning and Joint Optimization for Transformer-RNN-Transducer Speech Recognition.
Jae-Jin JeonEesung KimPublished in: ICASSP (2021)
Keyphrases
- joint optimization
- speech recognition
- multitask learning
- nearest neighbor
- gaussian processes
- multi task
- learning tasks
- multi task learning
- hidden markov models
- language model
- pattern recognition
- semi supervised
- sparse representation
- labeled data
- gaussian process
- speech signal
- training set
- speech recognition systems
- information retrieval
- semi supervised learning
- three dimensional