LRSpeech: Extremely Low-Resource Speech Synthesis and Recognition.
Jin XuXu TanYi RenTao QinJian LiSheng ZhaoTie-Yan LiuPublished in: CoRR (2020)
Keyphrases
- speech synthesis
- speech corpus
- speech recognition
- text to speech
- recognition rate
- pattern recognition
- low error rates
- vocal tract
- recognition accuracy
- automatic recognition
- recognition algorithm
- object recognition
- image recognition
- feature extraction
- prosodic features
- visual recognition
- high levels
- recognition process
- recognition scheme
- resource allocation
- handwriting recognition
- unsupervised learning
- image matching
- probabilistic model
- hidden markov models
- resource management
- human activities