SuperVLAD: Compact and Robust Image Descriptors for Visual Place Recognition | Read Paper on Bytez