Broadband doa estimation using convolutional neural networks trained with noise signals.
Soumitro ChakrabartyEmanuël A. P. HabetsPublished in: WASPAA (2017)
Keyphrases
- convolutional neural networks
- convolutional network
- doa estimation
- direction of arrival
- signal to noise ratio
- signal subspace
- canonical correlation analysis
- pattern recognition
- coarse to fine
- noise level
- training set
- computer simulation
- sound source
- speech recognition
- augmented reality
- independent component analysis