Return to Article Details A bilingual analysis of multi-head attention mechanism for image captioning based on morphosyntactic information Download Download PDF