Is Similarity Visually Grounded? Computational Model of Similarity for the Estonian language

19753

Knowledge Triples