List your company

Image Captioning Evaluation on MCEval 1.0 (test)

87.8Real Style ScoreDISCODE
56.70464.77772.8580.923Dec 16, 2025

Evaluation results

MethodLinks
DISCODELVLM=true, finetuning-free=true, backbone=LLaVA-Next-72B, Evaluation Category=Reference-free2025.1287.885.283.978.583.582.883.6
FLEUR (our implementation)LVLM=true, finetuning-free=true, backbone=LLaVA-Next-72B, Evaluation Category=Reference-free2025.1286.984.383.176.38282.382.5
G-VEval (GPT-4o)LVLM=true, finetuning-free=true, Evaluation Category=Reference-free2025.128680.281.276.980.68181
FLEURLVLM=true, finetuning-free=true, Evaluation Category=Reference-free2025.1284.783.680.445.679.98676.7
PolosLVLM=false, finetuning-free=false, Evaluation Category=Reference-based2025.1281.37577.676.874.56975.7
PAC-SLVLM=false, finetuning-free=false, Evaluation Category=Reference-free2025.1280.771.169.767.566.858.769.1
CLIP-SLVLM=true, finetuning-free=false, Evaluation Category=Reference-free2025.1279.27878.375.473.966.775.3
BERT-SLVLM=true, finetuning-free=false, Evaluation Category=Reference-based2025.1268.573.874.372.66758.669.1
METEORLVLM=false, finetuning-free=false, Evaluation Category=Reference-based2025.1267.864.867.561.860.859.463.7
CIDErLVLM=false, finetuning-free=false, Evaluation Category=Reference-based2025.1266.764.568.762.864.560.264.6
BLEU-4LVLM=false, finetuning-free=false, Evaluation Category=Reference-based2025.1259.358.161.555.754.953.257.1
ROUGELVLM=false, finetuning-free=false, Evaluation Category=Reference-based2025.1257.956.959.554.554.450.655.6