Beyond One and Two Tower: Cross-Modal Consensus Learning for Image-Text Retrieval | IEEE Journals & Magazine | IEEE Xplore