Reconstructing Vechicles from a Single Image: Shape Priors for Road Scene Understanding | Read Paper on Bytez