Multi-Scale Spectrogram Modelling for Neural Text-to-Speech.
Ammar AbbasBajibabu BollepalliAlexis MoinetArnaud JolyPenny KaranasouPeter MakarovSimon SlangenSri KarlapatiThomas DrugmanPublished in: CoRR (2021)
Keyphrases
- text to speech
- multiscale
- speech synthesis
- neural network
- network architecture
- prosodic features
- programming tool
- text to speech synthesis
- image processing
- english text
- coarse to fine
- scale space
- wavelet transform
- image segmentation
- keypoints
- pattern analysis
- image representation
- neural model
- natural images
- pattern recognition
- biologically inspired
- speech signal
- learning rules
- multiple scales
- edge detection