List your company

Text Retrieval on MS-COCO 5K (test)

84.8R@1CoCa
69.40873.40477.481.396Apr 13, 2020Nov 9, 2020Jun 8, 2021Jan 4, 2022Aug 3, 2022Mar 1, 2023Sep 28, 2023

Evaluation results

MethodLinks
CoCaParameters=2.1B, Evaluation Protocol=Fine-Tuned, Training Data=Super-Large2023.0184.896.598.3
X-FMbaseParameters=327M, Evaluation Protocol=Fine-Tuned, Training Data=More Data2023.0184.296.498.4
X2-VLMbaseParameters=255M, Evaluation Protocol=Fine-Tuned, Training Data=More Data2023.0183.596.398.5
VLMO-Large++Pretrain Images=1.0B, Model Size Category=Large-Size, Interaction Protocol=Dual encoder (dot product), large batch size=true2021.1183.19698.2
mPLUG-2#PT Data=17M2023.0282.595.798
BLIP#PT Data=129M2023.0282.495.497.9
BLIP2023.0582.495.497.9
OmniVL# Img-Text Pairs=14M*, Fine-tuned=true2022.0982.195.998.1
BLIPPre-training data size=129M2022.0681.995.497.8
Florence-HugePretrain Images=900M, Model Size Category=Huge-Size, Interaction Protocol=Dual encoder (dot product)2021.1181.895.2—
Florence#PT Data=0.9B2023.0281.895.2—
X-FMbaseParameters=327M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.0181.89698.3
Florence# Img-Text Pairs=900M, Fine-tuned=true2022.0981.895.2—
BEIT-3Parameters=1.9B, Evaluation Protocol=Zero-Shot, Training Data=Super-Large2023.0181.595.697.8
X-VLM# Params=216M, # Pre-train Images=16M2021.1181.295.698.2
mPLUG-2Base#PT Data=17M2023.0281.295.698.1
X-VLM + EPICData=4M+, EPIC=true, Protocol=Fine-tuned2022.118195.397.9
X-VLM + EPICData=16M+, EPIC=true, Protocol=Fine-tuned2022.1180.795.698
BLIPPre-training data size=14M2022.0680.695.297.6
BLIP# Img-Text Pairs=14M, Fine-tuned=true2022.0980.695.297.6
X2-VLMbaseParameters=255M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.0180.595.597.8
X-VLMParameters=216M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.0180.495.598.2
X-VLM# Params=216M, # Pre-train Images=4M2021.1180.195.598.2
X-VLMData=4M+, EPIC=false, Protocol=Fine-tuned2022.1179.895.197.7
METER + EPICData=16M, EPIC=true, Protocol=Fine-tuned2022.1179.794.897.5
X-VLMData=16M+, EPIC=false, Protocol=Fine-tuned2022.1179.595.497.8
VL-BEITParameters=175M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.0179.5——
ALBEF + EPICData=16M, EPIC=true, Protocol=Fine-tuned2022.1179.294.797.5
METER + EPICData=4M, EPIC=true, Protocol=Fine-tuned2022.117994.597.5
ALBEFData=16M, EPIC=false, Protocol=Fine-tuned2022.1178.393.996.8
VLMO-LargePretrain Images=4M, Model Size Category=Large-Size, Interaction Protocol=Dual encoder (dot product)2021.1178.294.497.4
VLMo#PT Data=4M2023.0278.294.497.4
METERData=16M, EPIC=false, Protocol=Fine-tuned2022.1178.293.896.9
CCLM_basePre-training Data=4M2022.0677.72——
ALBEFPre-train Images=14M, Fine-tuned=true2021.0777.694.397.2
ALBEFPre-training data size=14M2022.0677.694.397.2
ALBEF# Params=210M, # Pre-train Images=14M2021.1177.694.397.2
ALBEF#PT Data=14M2023.0277.694.397.2
X-FMbaseParameters=327M, Evaluation Protocol=Zero-Shot, Training Data=More Data2023.0177.694.897.7
ALBEF# Img-Text Pairs=14M, Fine-tuned=true2022.0977.694.397.2
RaSa2023.0577.4494.1297.18
AlignCMSS2023.0977.494.197
ALBEFconfiguration=backbone2023.0577.2694.0297.04
METERData=4M, EPIC=false, Protocol=Fine-tuned2022.1177.293.797.1
ALIGNPre-train Images=1.2B, Fine-tuned=true2021.077793.596.9
ALIGN-LargePretrain Images=1.8B, Model Size Category=Large-Size, Interaction Protocol=Dual encoder (dot product)2021.117793.596.9
ALIGNPre-training data size=1.2B2022.067793.596.9
SINGULARITYPre-training data size=17M2022.067793.796.8
ALIGN# Params=490M, # Pre-train Images=1.8B2021.117793.596.9
ALIGNTraining Data=1.8B2022.077793.596.9
ALIGN#PT Data=1.8B2023.027793.596.9
ALIGN2023.057793.596.9
ALIGN# Img-Text Pairs=1.8B, Fine-tuned=true2022.097793.596.9
OmniVLParameters=288M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.0176.893.697.3
OmniVL# Img-Text Pairs=4M*, Fine-tuned=true2022.0976.893.697.3
METER-CLIP# Params=380M, # Pre-train Images=4M2021.1176.293.296.8
METER# Img-Text Pairs=404M, Fine-tuned=true2022.0976.293.296.8
OSCAR+ w/ VINVLBERT scale=Large2021.0175.492.996.2
VinVL-LargePretrain Images=5.7M, Model Size Category=Large-Size, Interaction Protocol=Fusion encoder2021.1175.492.996.2
VinVL_large# Params=550M, # Pre-train Images=5.6M2021.1175.492.996.2
ALBEF + EPICData=4M, EPIC=true, Protocol=Fine-tuned2022.1175.192.996.4
VLMO-BasePretrain Images=4M, Model Size Category=Base-Size, Interaction Protocol=Dual encoder (dot product)2021.1174.893.196.9
VLMbaseParameters=175M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.0174.893.196.9
VLMO# Img-Text Pairs=4M, Fine-tuned=true2022.0974.893.196.9
OSCAR+ w/ VINVLBERT scale=Base2021.0174.692.696.3
VinVL-BaseVisual Embed=Region, Time (ms)=~650, Extra pre-training=GQA, VQAv2, VG-QA, Extra pre-training dataset=Open Images2021.0274.692.696.3
VinVL (Base)Training Data=8.9M2022.0774.692.696.3
VinVL_baseModel Scale=base2022.0674.6——
VinVL2023.0974.692.696.3
X-FMbaseParameters=327M, Evaluation Protocol=Zero-Shot, Training Data=Standard2023.0173.893.997.2
OSCARModel Size=Large2020.0473.592.296
OSCARSize=L2020.0473.592.296
OscarEvaluation Protocol=Finetune2021.0273.592.296
OscarBERT scale=Large2021.0173.592.296
OSCARMode=Fine-tuned2022.1073.592.296
Oscar2023.0573.592.296
ALBEFData=4M, EPIC=false, Protocol=Fine-tuned2022.1173.392.496.4
X-VLMData=16M+, EPIC=true, Evaluation Protocol=zero-shot2022.1173.293.696.9
X-VLM + EPICData=16M+, EPIC=true, Protocol=Zero-shot2022.1173.293.696.9
ALBEFPre-train Images=4M, Fine-tuned=true2021.0773.191.496
ALBEF-BasePretrain Images=4M, Model Size Category=Base-Size, Interaction Protocol=Fusion reranking2021.1173.191.496
ALBEFPre-training data size=4M2022.0673.191.496
ALBEF# Params=210M, # Pre-train Images=4M2021.1173.191.496
ALBEFPre-training Data=4M2022.0673.1——
X-VLMData=16M+, EPIC=false, Evaluation Protocol=zero-shot2022.1173.192.897
X-VLMData=16M+, EPIC=false, Protocol=Zero-shot2022.1173.192.897
ALBEFParameters=210M, Evaluation Protocol=Fine-Tuned, Training Data=Standard2023.0173.191.496
METER-Swin# Params=380M, # Pre-train Images=4M2021.11739296.3
X-VLMData=4M+, EPIC=true, Evaluation Protocol=zero-shot2022.117293.497.3
X-VLM + EPICData=4M+, EPIC=true, Protocol=Zero-shot2022.117293.497.3
SINGULARITYPre-training data size=5M2022.0671.990.895.4
X2-VLMbaseParameters=255M, Evaluation Protocol=Zero-Shot, Training Data=More Data2023.0171.793.497.5
X-VLM# Params=216M, # Pre-train Images=16M, Zero-shot=true2021.1171.693.197
COOKIE2023.0571.690.995.4
X-VLM# Params=216M, # Pre-train Images=4M, Zero-shot=true2021.1170.892.196.5
X-VLMParameters=216M, Evaluation Protocol=Zero-Shot, Training Data=Standard2023.0170.892.196.5
Knowledge-CLIPMode=Fine-tuned2022.1070.289.294.4
OSCARModel Size=Base2020.047091.195.5
OSCARSize=B2020.047091.195.5
OscarBERT scale=Base2021.017091.195.5