Enabling On-Device Training of Speech Recognition Models With Federated Dropout.
Dhruv GulianiLillian ZhouChangwan RyuTien-Ju YangHarry ZhangYonghui XiaoFrançoise BeaufaysGiovanni MottaPublished in: ICASSP (2022)
Keyphrases
- speech recognition
- acoustic models
- speech signal
- automatic speech recognition
- speech synthesis
- speech recognizer
- hidden markov models
- speech processing
- language model
- isolated word
- probabilistic model
- pattern recognition
- speech recognition technology
- noisy environments
- speech recognition systems
- wall street journal corpus
- speech recognition errors
- speech enhancement
- discriminative training