The project deals with the problem of automatically assessing the characteristics of images that refer to a particular text. For example, we want to assess which of the main objects mentioned in the text are present in the images. Or how aspects, that are left open in the text, are visualized in the images.