Efficient methods to train multilingual bottleneck feature extractors for low resource keyword search

• Published in: 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 5650 - 5654
• Main Authors: Chongjia Ni, Cheung-Chi Leung, Lei Wang, Chen, Nancy F., Bin Ma
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.03.2017
• Subjects: Buildings; Data mining; Feature extraction; Keyword spotting; language identification; multilingual data selection; recurrent neural network; Recurrent neural networks; Speech; Training; Training data
• ISSN: 2379-190X
• DOI: 10.1109/ICASSP.2017.7953238

Training a bottleneck feature (BNF) extractor with multilingual data has been common in low resource keyword search. In a low resource application, the amount of transcribed target language data is limited while there are usually plenty of multilingual data. In this paper, we investigated two methods to train efficient multilingual BNF extractors for low resource keyword search. One method is to use the target language data to update an existing BNF extractor, and another method is to combine the target language data to train a new multilingual BNF extractor from the start. In these two methods, we proposed to use long short-term memory recurrent neural network based language identification to select utterances in the multilingual training data that are acoustically close to the target language. Experiments on Swahili in the OpenKWS15 data demonstrated the efficiency of our proposed methods. The first method facilitates rapid system development, while both methods outperform using baseline BNF extractors in terms of accuracy.