ANÁLISE DAS DESCRIÇÕES DE IMAGENS GERADAS POR INTELIGÊNCIA ARTIFICIAL APLICADAS EM REDES SOCIAIS

Abstract

Artificial Intelligence (AI) is increasingly present in our daily lives. With the frequent use of social media by people with visual impairments, platforms such as Facebook and Instagram have begun to use AI to generate automatic descriptions of images for this audience. However, can AI provide satisfactory descriptions that convey all the information contained in the images published on social media? This study aims to analyze the descriptions of images generated by AI and human descriptions on the social media platforms Facebook and Instagram, from the perspective of blind users. To evaluate the effectiveness of this resource, a sample test was conducted with seven legally blind or visually impaired participants, using interviews and tests on both social networks. The results of the research indicated that the descriptions generated by AI were succinct and did not cover all the details of the images tested when compared to the descriptions provided by humans. The data obtained showed that, although succinct descriptions provided by AI are more useful than the absence of any description, Artificial Intelligence played a significant role in connecting more blind users on social media. However, continuous improvement is needed to promote more effective inclusion for all users.