Online Keyword Spotting with a Character-Level Recurrent Neural Network

@article{Hwang2015OnlineKS,
  title={Online Keyword Spotting with a Character-Level Recurrent Neural Network},
  author={Kyuyeon Hwang and Minjae Lee and Wonyong Sung},
  journal={CoRR},
  year={2015},
  volume={abs/1512.08903}
}
In this paper, we propose a context-aware keyword spotting m odel employing a character-level recurrent neural network (RNN) for spoken term detection in co tinuous speech. The RNN is end-toend trained with connectionist temporal classification (CT C) to generate the probabilities of character and word-boundary labels. There is no need for the phonetic t ranscription, senone modeling, or system dictionary in training and testing. Also, keywords can easi ly be added and modified by editing the… CONTINUE READING
Related Discussions
This paper has been referenced on Twitter 2 times. VIEW TWEETS

References

Publications referenced by this paper.
Showing 1-10 of 18 references

Long Short-Term Memory

Neural Computation • 1997
View 9 Excerpts
Highly Influenced

A hidden Markov model based keyw ord recognition system

R. C. Rose, D. B. Paul
inAcoustics, Speech, and Signal Processing, 1990. ICASSP-90., 1990 International Confere ce on. IEEE, 1990, pp. 129–132. • 1990
View 4 Excerpts
Highly Influenced

Single stream parallelization of generalized LSTM-like RNNs on a GPU

2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) • 2015
View 1 Excerpt

Small-footprint keyword spotting using deep neural networks

2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) • 2014
View 1 Excerpt

Similar Papers

Loading similar papers…