Hello,
I would like to ask a question about ROUGE-1 metric score. For the task of caption prediction, which of 3 measures will you take into consideration for ROUGE-1? Accuracy, Recall or F-measure?
Thanks in advance,
Georgios Zachariadis
Hello,
I would like to ask a question about ROUGE-1 metric score. For the task of caption prediction, which of 3 measures will you take into consideration for ROUGE-1? Accuracy, Recall or F-measure?
Thanks in advance,
Georgios Zachariadis
Hello Georgios,
we are using the F-measure.
Kind regards,
Johannes
@saviola777 But according to ImageCLEFmed Caption | ImageCLEF / LifeCLEF - Multimedia Retrieval in CLEF , BLEU and ROUGE scores are supposed to be used for caption prediction??
Yes, that is correct. BLEU is used as the primary score, ROUGE is used as the secondary score.