Semantic Quantization of 3D Human Motion Capture Data Through Spatial-Temporal Feature Extraction