Evaluation of Google Image Translate in Rendering Arabic Signage into English

When people travel to another country for work or leisure, they regularly need a medium help them understand the written messages in other languages. Google Translate offers new service: translating content of images (texts) instantly and freely into 100 languages powered by Neural Machine Translation approach (NMT). In this vein, current research paper attempts evaluate accuracy Image service rendering texts printed on Arabic signage: banners road shop signs from English. Besides, it aims identify capacity signage English effectively without human translators. The adopts Linguistic Error Analysis Framework Costa et al. (2015) analyzing output image terms orthography, grammar, lexis, semantics. shows that made following errors while English: mistranslation, omission, additions, wrong choice, misordering, subject-verb disagreement, semantic errors. conclusion, helps users configure gist image. However, translator is still needed since MT may not provide an adequate effective translation as humans do.