论文笔记-SPEAKER RECOGNITION FROM RAW WAVEFORM WITH SINCNET
DNNshavealsobeenproposedfordirectdiscrimnativespeakerclassification,aswitnessedbytherecentliteratureonthistopic[13–16].Mostofpastattempts,however,employedhand-craftedfeaturessuchasFBANKandMFCCcoeffici