上一条： LSTM-Based Pitch Range Estimation from Spectral Information of Brief Speech Input

下一条： A Multi-modal Soft Targets Approach for Pronunciation Erroneous Tendency Detection