Prediction of intelligibility of noisy and time-frequency weighted speech based on mutual information between amplitude envelopes.
Jesper JensenCees H. TaalPublished in: INTERSPEECH (2013)
Keyphrases
- mutual information
- speech recognition
- similarity measure
- prediction accuracy
- noisy environments
- image registration
- information theoretic
- medical image registration
- linear prediction
- feature selection
- prediction model
- prediction error
- information gain
- speech synthesis
- prediction algorithm
- speech signal
- signal processing
- signal to noise ratio
- sum of squared differences
- conditional mutual information
- audio visual
- neural network
- noise free
- noise reduction
- pattern recognition
- instantaneous frequency