Feature enhancement by deep LSTM networks for ASR in reverberant multisource environments.
Felix WeningerJürgen T. GeigerMartin WöllmerBjörn W. SchullerGerhard RigollPublished in: Comput. Speech Lang. (2014)
Keyphrases
- multi source
- information fusion
- data fusion
- single source
- speech signal
- automatic speech recognition
- social networks
- complex networks
- real world
- data integration
- fusion model
- community detection
- speech recognition
- feature set
- database systems
- decision making
- artificial intelligence
- recurrent neural networks
- data sources
- case study
- machine learning
- neural network