Combining De-noising Auto-encoder and Recurrent Neural Networks in End-to-End Automatic Speech Recognition for Noise Robustness.
Tzu-Hsuan TingChia-Ping ChenPublished in: SLT (2018)
Keyphrases
- end to end
- recurrent neural networks
- automatic speech recognition
- noisy environments
- denoising
- speech recognition
- speech signal
- neural network
- broadcast news
- hidden markov models
- rate allocation
- feed forward
- recurrent networks
- artificial neural networks
- echo state networks
- congestion control
- noise level
- noisy images
- speech retrieval
- noise reduction
- bit rate
- conversational speech
- signal to noise ratio
- rate distortion
- noise model
- scalable video
- motion estimation
- real time
- video codec
- low complexity
- machine learning