The challenge of automatically describing images and videos has stimulated much research in Computer Vision and Natural Language Processing. In order to test the semantic abilities of new algorithms, we need reliable and objective ways of measuring progress. Using our dataset of 2K human and machine descriptions, we find that standard evaluation measures alone do not adequately measure the semantic richness of a description. We introduce and test a new measure of semantic ability based on relative lexical diversity. We show how our measure can work alongside existing measures to achieve state of the art correlation with human judgement of quality.
展开▼