Speaker Identification Based on Efficiently Combining Manifold Features.

• 基于多特征有效组合的说话人识别。