Dilated Temporal Fully-Convolutional Network for Semantic Segmentation of Motion Capture Data | Read Paper on Bytez