Sciweavers

Share
ICASSP
2009
IEEE

Robust discriminative keyword spotting for emotionally colored spontaneous speech using bidirectional LSTM networks

10 years 4 months ago
Robust discriminative keyword spotting for emotionally colored spontaneous speech using bidirectional LSTM networks
In this paper we propose a new technique for robust keyword spotting that uses bidirectional Long Short-Term Memory (BLSTM) recurrent neural nets to incorporate contextual information in speech decoding. Our approach overcomes the drawbacks of generative HMM modeling by applying a discriminative learning procedure -linearly maps speech features into an abstract vector space. By incorporating the outputs of a BLSTM network into the speech features, it is able to make use of past and future context for phoneme predictions. The robustness of the approach is evaluated on a keyword spotting task using the HUMAINE Sensitive Artificial Listener (SAL) database, which contains accented, spontaneous, and emotionally colored speech. The test is particularly stringent because the system is not trained on the SAL database, but only on the TIMIT corpus of read speech. We show that our method prevails over a discriminative keyword spotter without BLSTM-enhanced feature functions, which in turn has ...
Martin Wöllmer, Florian Eyben, Joseph Keshet,
Added 21 May 2010
Updated 21 May 2010
Type Conference
Year 2009
Where ICASSP
Authors Martin Wöllmer, Florian Eyben, Joseph Keshet, Alex Graves, Björn Schuller, Gerhard Rigoll
Comments (0)
books