Phoneme recognition in TIMIT with BLSTM-CTC

TL;DR

This paper demonstrates that a single BLSTM-CTC neural network can achieve phoneme recognition performance on TIMIT comparable to complex classifier combinations, simplifying the approach.

Contribution

The study shows that a single recurrent neural network with BLSTM-CTC can match the performance of more complex classifier ensembles on TIMIT phoneme recognition.

Findings

Achieved 24.6% error rate on TIMIT phoneme recognition.

Single BLSTM-CTC network performs comparably to classifier combinations.

Simplifies phoneme recognition approach without sacrificing accuracy.

Abstract

We compare the performance of a recurrent neural network with the best results published so far on phoneme recognition in the TIMIT database. These published results have been obtained with a combination of classifiers. However, in this paper we apply a single recurrent neural network to the same task. Our recurrent neural network attains an error rate of 24.6%. This result is not significantly different from that obtained by the other best methods, but they rely on a combination of classifiers for achieving comparable performance.