A Multiresolution-Based Fusion Strategy for Improving Speech Emotion Recognition Efficiency.