Deep neural network features and semi-supervised training for low resource speech recognition

We propose a new technique for training deep neural networks (DNNs) as data-driven feature front-ends for large vocabulary con-tinuous speech recognition (LVCSR) in low resource settings. To circumvent the lack of sufficient training data for acoustic mod-eling in these scenarios, we use transcribed multilingual data and semi-supervised training to build the proposed feature front-ends. In our experiments, the proposed features provide an absolute im-provement of 16 % in a low-resource LVCSR setting with only one hour of in-domain training data. While close to three-fourths of these gains come from DNN-based features, the remaining are from semi-supervised training. Index Terms — Low resource, speech recognition, deep neural networks, semi-supervised training, bottleneck features