| Popis: |
Recently, many tourists visit Japan from various countries. Hence, automatic speech translation device is desired for services in various business scenes at a shopping, a station, a town, and a hotel. For realizing speech translation with higher accuracy under these situations, speech from user must be captured and recognized accurately even under various kinds of noise environments. In this study, we developed adaptive beamformer technology for speech enhancement under noisy environments and evaluated speech recognition performance. |
