CLIP-Decoder : ZeroShot Multilabel Classification using Multimodal CLIP Aligned Representations | IEEE Conference Publication | IEEE Xplore