Semantic Segmentation for complete scene understanding

Developing a semantic segmentation model to classify each pixel among the everyday traffic scene objects. Exploring advanced CNN algorithms: Xception, Inception-v4, ResNeXt-50, U-Net.