System-level Correlation with Human Judgment on MS COCO M1 (system-level) benchmark leaderboard · SOTA2 Research

System-level Correlation with Human Judgment on MS COCO M1 (system-level)

0.986Pearson rBERTScore

Evaluation Results

Method	Links
BERTScore2021.06		0.986	0.014
SMURF2021.06		0.984	0.016
SPICE2021.06		0.956	0.044
SPURTS2021.06		0.956	0.044
SPARCS2021.06		0.874	0.126
METEOR2021.06		0.479	0.521
BS-w/oidf2021.06		0.374	0.626
CIDEr2021.06		0.023	0.977
Bleu-12021.06		-0.279	0.721
Bleu-22021.06		-0.709	0.291
Rouge-L2021.06		-0.812	0.188