Self-Supervised Learning of Intertwined Content and Positional Features for Object Detection | Read Paper on Bytez