Multi-Source Direction of Arrival Estimation of Noisy Speech using Convolutional Recurrent Neural Networks with Higher-Order Ambisonics Signals.
Nils PoschadelStephan PreihsJürgen PeissigPublished in: EUSIPCO (2021)
Keyphrases
- recurrent neural networks
- multi source
- direction of arrival
- doa estimation
- higher order
- signal subspace
- data fusion
- sensor array
- information fusion
- canonical correlation analysis
- multiple sources
- neural network
- feed forward
- artificial neural networks
- data sources
- data integration
- natural images
- signal processing
- markov random field
- pairwise
- correlation matrix
- signal to noise ratio
- speech recognition
- sound source
- feature selection
- data sets
- community detection
- estimation error
- noisy environments
- data mining