English

Utterance-level end-to-end language identification using attention-based CNN-BLSTM

Audio and Speech Processing 2019-02-21 v1 Machine Learning Sound

Abstract

In this paper, we present an end-to-end language identification framework, the attention-based Convolutional Neural Network-Bidirectional Long-short Term Memory (CNN-BLSTM). The model is performed on the utterance level, which means the utterance-level decision can be directly obtained from the output of the neural network. To handle speech utterances with entire arbitrary and potentially long duration, we combine CNN-BLSTM model with a self-attentive pooling layer together. The front-end CNN-BLSTM module plays a role as local pattern extractor for the variable-length inputs, and the following self-attentive pooling layer is built on top to get the fixed-dimensional utterance-level representation. We conducted experiments on NIST LRE07 closed-set task, and the results reveal that the proposed attention-based CNN-BLSTM model achieves comparable error reduction with other state-of-the-art utterance-level neural network approaches for all 3 seconds, 10 seconds, 30 seconds duration tasks.

Keywords

Cite

@article{arxiv.1902.07374,
  title  = {Utterance-level end-to-end language identification using attention-based CNN-BLSTM},
  author = {Weicheng Cai and Danwei Cai and Shen Huang and Ming Li},
  journal= {arXiv preprint arXiv:1902.07374},
  year   = {2019}
}

Comments

Accepted for ICASSP 2019

R2 v1 2026-06-23T07:45:36.835Z