Assessing multilingual multimodal image description: Studies of native speaker preferences and translator choices. (23rd April 2018)

• Record Type: Journal Article
• Authors: FRANK, STELLA; ELLIOTT, DESMOND; SPECIA, LUCIA
• Editors: Belz, Anja; Berg, Tamara
• Is Part Of: Natural language engineering. Volume 24:Part 3(2018)
• Page Start: 393
• Page End: 413
• Publication Date: 2018-04-23
• Subjects: Natural language processing (Computer science) -- Periodicals; Software engineering -- Periodicals; 006.35
• DOI: 10.1017/S1351324918000074
• Languages: English
• ISSNs: 1351-3249

Abstract

Abstract: Two studies on multilingual multimodal image description provide empirical evidence towards two questions at the core of the task: (i) whether target language speakers prefer descriptions generated directly in their native language, as compared to descriptions translated from a different language; (ii) whether images improve human translation of descriptions. These results provide guidance for future work in multimodal natural language processing by first showing that on the whole, translations are not distinguished from native language descriptions, and second delineating and quantifying the information gained from the image during the human translation task.