Long, Deep and Wide Artificial Neural Nets for Dealing with Unexpected Noise in Machine Recognition of Speech.
Hynek HermanskyPublished in: TSD (2013)
Keyphrases
- neural nets
- noisy environments
- recognition engine
- feed forward
- noisy speech
- speech recognition
- recognition rate
- back propagation
- automatic speech recognition systems
- neural network
- object recognition
- artificial neural networks
- speech enhancement
- supervised training
- speech signal
- pattern recognition
- noise reduction
- speech corpus
- learning tasks
- signal to noise ratio
- feature extraction
- real world
- recognition process
- hidden markov models
- speaker identification
- speech synthesis
- data mining
- unsupervised learning
- counter propagation
- deep architectures
- decision making
- speech recognition systems
- single layer
- image processing
- fuzzy logic