Interpreting intermediate convolutional layers of CNNs trained on raw speech.
Gasper BegusAlan ZhouPublished in: CoRR (2021)
Keyphrases
- convolutional network
- deep belief networks
- speech recognition
- cellular neural networks
- convolutional neural networks
- raw data
- training set
- multi layer
- audio visual
- speech signal
- isolated word
- text to speech
- spontaneous speech
- high level
- speaker recognition
- restricted boltzmann machine
- speech synthesis
- learning algorithm
- artificial neural networks
- feature selection
- multiple layers
- speaker identification
- spoken language
- genetic algorithm
- language acquisition
- sparse coding