VisualWord2Vec (Vis-W2V): Learning Visually Grounded Word Embeddings Using Abstract Scenes | IEEE Conference Publication | IEEE Xplore