Dynamic Layer Customization for Noise Robust Speech Emotion Recognition in Heterogeneous Condition Training.
Alex WilfEmily Mower ProvostPublished in: CoRR (2020)
Keyphrases
- speech emotion recognition
- image noise
- noise level
- salt pepper
- noisy environments
- dynamic environments
- sufficient conditions
- multi layer
- robust estimation
- imaging artifacts
- greater robustness
- highly dynamic
- neural network
- estimation error
- mobile robot
- multiscale
- real world
- training algorithm
- noise model
- noise reduction
- computationally efficient
- feedforward neural networks
- registration errors
- random noise
- application layer
- robust statistical
- user interface
- training set