Gradations of error severity in Automatic Image Descriptions in English
dc.rights.license | CC-BY-NC-ND | |
dc.contributor.advisor | Chen, G. | |
dc.contributor.author | Heracleous, Tasos | |
dc.date.accessioned | 2022-09-09T02:02:40Z | |
dc.date.available | 2022-09-09T02:02:40Z | |
dc.date.issued | 2022 | |
dc.identifier.uri | https://studenttheses.uu.nl/handle/20.500.12932/42546 | |
dc.description.sponsorship | Utrecht University | |
dc.language.iso | EN | |
dc.subject | Following a research carried by Miltenburg et al. (2020) we performed an experiment on English speakers to validate if different kinds of errors in image descriptions, elicit different evaluation scores. Our results show that the severity of different kinds of errors is perceived differently by humans which give different evaluation scores to each error type according either solely to the text they have read or the picture they have seen. | |
dc.title | Gradations of error severity in Automatic Image Descriptions in English | |
dc.type.content | Master Thesis | |
dc.rights.accessrights | Open Access | |
dc.subject.courseuu | Artificial Intelligence | |
dc.thesis.id | 9590 |