List your company

Image Captioning on nocaps standard (test)

124.8CIDErGIT2
4.57635.7886798.212May 27, 2022Nov 13, 2022May 3, 2023Oct 21, 2023Apr 9, 2024Sep 27, 2024Mar 17, 2025

Evaluation results

MethodLinks
GIT22022.05124.8
GIT2022.05123.4
GIT2022.05123.4
CoCa2022.05120.6
Prior SOTA2022.05120.6
UniversalCap2022.05119.3
SimVLM2022.05115.2
LEMON2022.05114.3
VinVL2022.0592.5
UFO2022.0592.3
VIVO2022.0586.6
Human2022.0585.3
Regular DecodingModel=LLaVA-v1.5-13B2025.0382.6
VAFModel=LLaVA-v1.5-13B2025.0382.3
OSCAR2022.0580.9
VAFModel=LLaVA-v1.5-7B2025.0378.8
Regular DecodingModel=LLaVA-v1.5-7B2025.0378.7
VCDModel=LLaVA-v1.5-13B2025.0368.9
VCDModel=LLaVA-v1.5-7B2025.0365.7
ICDModel=LLaVA-v1.5-7B2025.0362.3
ICDModel=LLaVA-v1.5-13B2025.0360.3
BLIP-2 (129M)Protocol=Zero-Shot Transfer, Specific Modules=ViT-G, Qformer, Examples Num=4, #Params=12.1B2024.0315.8
GiT-HTraining=universal, Specific Modules=None, Examples Num=1, #Params=756M2024.0312.6
GiT-LTraining=universal, Specific Modules=None, Examples Num=1, #Params=387M2024.0311.6
GiT-BTraining=universal, Specific Modules=None, Examples Num=1, #Params=131M2024.0310.6
GiT-BTraining=multi-task, Specific Modules=None, Examples Num=1, #Params=131M2024.039.2