Improving LSTM-CTC based ASR performance in domains with limited training data

@article{Billa2017ImprovingLB,
  title={Improving LSTM-CTC based ASR performance in domains with limited training data},
  author={Jayadev Billa},
  journal={CoRR},
  year={2017},
  volume={abs/1707.00722}
}
This paper addresses the observed performance gap between automatic speech recognition (ASR) systems based on Long Short Term Memory (LSTM) neural networks trained with the connectionist temporal classification (CTC) loss function and systems based on hybrid Deep Neural Networks (DNNs) trained with the cross entropy (CE) loss function on domains with limited data. We step through a number of experiments that show incremental improvements on a baseline EESEN toolkit based LSTM-CTC ASR system… CONTINUE READING
Related Discussions
This paper has been referenced on Twitter 18 times. VIEW TWEETS

Citations

Publications citing this paper.

References

Publications referenced by this paper.
Showing 1-10 of 34 references

RNNDROP: A novel dropout for RNNS in ASR

2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU) • 2015
View 13 Excerpts
Highly Influenced

An empirical exploration of CTC acoustic models

2016 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) • 2016

Gomez , and Jürgen Schmidhuber . Connectionist temporal classification : labelling unsegmented sequence data with recurrent neural networks

K Kuldip
Machine Learning , Proceedings of the Twenty - Third International Conference ( ICML 2006 ) 2015 IEEE Workshop on Automatic Speech Recognition and Understanding , ASRU 2015 2015 IEEE International Conference on Acoustics , Speech and Signal Processing • 2016

Similar Papers

Loading similar papers…