Attenuation Of Acoustic Early Reflections In Television Studios Using Pretrained Speech Synthesis Neural Network.
Tomer RosenbaumIsrael CohenEmil WinebrandPublished in: ICASSP (2022)
Keyphrases
- speech synthesis
- neural network
- prosodic features
- speech recognition
- text to speech
- vocal tract
- pattern recognition
- back propagation
- genetic algorithm
- fuzzy logic
- associative memory
- self organizing maps
- activation function
- neural network model
- artificial neural networks
- auto associative
- speech corpus
- spatially varying
- network architecture
- multi layer
- speech signal
- network model
- feed forward neural networks
- bp neural network
- fuzzy artmap
- prediction model
- wave propagation
- noisy images
- underwater vehicles
- neural network is trained
- face recognition
- computer vision