1 paper
Daniel A. P. Oliveira, Lourenço Teodoro, David Martins de Matos
Current image captioning systems lack the ability to link descriptive text to specific visual elements, making their outputs difficult to verify. While recent approaches offer some…