In automatic speech recognition, often little training data is available for specific challenging tasks, but training of state-of-the-art automatic speech recognition systems requires large amounts of annotated speech. To address this issue, we propose a two-staged approach to acoustic modeling that combines noise and reverberation data augmentation with transfer learning to robustly address challenges such as difficult acoustic recording conditions, spontaneous speech, and speech of elderly people. We evaluate our approach using the example of German oral history interviews, where a relative average reduction of the word error rate by 19.3% is achieved.
@article{arxiv.1908.06709,
title = {Two-Staged Acoustic Modeling Adaption for Robust Speech Recognition by the Example of German Oral History Interviews},
author = {Michael Gref and Christoph Schmidt and Sven Behnke and Joachim Köhler},
journal= {arXiv preprint arXiv:1908.06709},
year = {2019}
}
Comments
Accepted for IEEE International Conference on Multimedia and Expo (ICME), Shanghai, China, July 2019