Generating Diverse and Descriptive Image Captions Using Visual Paraphrases
Recommended citation: Lixin Liu, Jiajun Tang, Xiaojun Wan, and Zongming Guo. "Generating Diverse and Descriptive Image Captions Using Visual Paraphrases." ICCV, 2019.
This paper improves the diversity and descriptiveness of machine-generated image captions by utilizing visual paragraphs in existing captioning datasets.