Article details

Research area
Speech recognition

Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference on


Zoltan Tuske, Joel Pinto, Daniel Willett, Ralf Schluter

Investigation on cross- and multilingual mlp features under matched and mismatched acoustical conditions


In this paper, Multi Layer Perceptron (MLP) based multilingual bottleneckfeatures are investigated for acoustic modeling in three languages— German, French, and US English. We use a modifiedtraining algorithm to handle the multilingual training scenario withouthaving to explicitly map the phonemes to a common phonemeset. Furthermore, the cross-lingual portability of bottleneck featuresbetween the three languages are also investigated. Single passrecognition experiments on large vocabulary SMS dictation task indicatethat (1) multilingual bottleneck features yield significantlylower word error rates compared to standard MFCC features (2) multilingualbottleneck features are superior to monolingual bottleneckfeatures trained for the target language with limited training data,and (3) multilingual bottleneck features are beneficial in training acoustic models in a low resource language where only mismatchedtraining data is available — by exploiting the more matched trainingdata from other languages.

Read/download now