Learning to Annotate Clothes in Everyday Photos: Multi-modal, Multi-label, Multi-instance Approach | IEEE Conference Publication | IEEE Xplore