Representation Learning for Visual Object Tracking by Masked Appearance Transfer | Read Paper on Bytez