Persian speech synthesis using enhanced tacotron based on multi-resolution convolution layers and a convex optimization method.
Navid NaderiBabak NasersharifAmirhossein NikoofardPublished in: Multim. Tools Appl. (2022)
Keyphrases
- optimization method
- speech synthesis
- multiresolution
- speech recognition
- text to speech
- optimization algorithm
- optimization methods
- optimization process
- differential evolution
- simulated annealing
- global optimum
- genetic algorithm
- vocal tract
- evolutionary algorithm
- prosodic features
- particle swarm
- nonlinear optimization
- optimization procedure
- nelder mead simplex
- metaheuristic
- speech corpus
- quasi newton
- convex hull
- wavelet transform
- text classification
- middle layer
- neural network
- subband
- hierarchical representation
- global optimization
- pattern recognition
- language model
- image processing
- multiscale
- hybrid algorithm
- speech signal
- density estimation