Fetching the paper…

Visual Word2Vec (vis-w2v): Learning Visually Grounded Word Embeddings Using Abstract Scenes · Around