BLEU-4: 14.45METEOR: 17.53ROUGE-L: 24.28CIDEr: 127.37
Papers with Code · CC BY-SA 4.0
FS-MEVQA
The Few-Shot Multimodal Explanation for Visual Question Answering (FS-MEVQA) task aims to learn MEVQA from few training samples.
논문
BLEU-4: 67.91METEOR: 50.55ROUGE-L: 79.41CIDEr: 510.44
BLEU-4: 41.87METEOR: 34.61ROUGE-L: 55.90CIDEr: 276.14
BLEU-4: 45.51METEOR: 35.17ROUGE-L: 52.67CIDEr: 269.68
Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
BLEU-4: 24.30METEOR: 23.40ROUGE-L: 34.52CIDEr: 201.47
BLEU-4: 0.00METEOR: 4.37ROUGE-L: 23.23CIDEr: 0.89
BLEU-4: 9.17METEOR: 19.82ROUGE-L: 33.34CIDEr: 4.28