Boosting Noise Robustness of Acoustic Model via Deep Adversarial Training.
Bin LiuShuai NieYaping ZhangDengfeng KeShan LiangWenju LiuPublished in: CoRR (2018)
Keyphrases
- early stopping
- training process
- learning algorithm
- deep architectures
- boosting algorithms
- additive noise
- data sets
- signal to noise ratio
- training and testing data
- noise reduction
- training set
- greater robustness
- training phase
- noise immunity
- weighted sums
- image transformations
- unsupervised feature learning
- noisy environments
- noise model
- training algorithm
- ensemble learning
- ensemble methods
- training samples
- active learning
- feature selection
- neural network