Multiband, multisensor robust features for noisy speech recognition.
Dimitrios DimitriadisPetros MaragosStamatios LefkimmiatisPublished in: INTERSPEECH (2007)
Keyphrases
- speech recognition
- noisy environments
- multi sensor
- multi band
- data fusion
- speech recognition systems
- hidden markov models
- speech signal
- feature extraction
- automatic speech recognition
- language model
- image features
- low level
- infrared
- classification accuracy
- pattern recognition
- speaker identification
- neural network
- feature set
- feature space
- frequency band
- high quality
- high level