说明：  音节中心音的检测。输入为帧级别的特征，每一帧有多于12维特征；检测十二平均律在该首音乐上的中心音。
(Tone syllable center detection. Input for the frame-level features, each frame has more than 12-dimensional features detection十二平均律in the center of the first music on the sound.)