List your company

Text-to-Video Retrieval on MSRVTT 1k (test)

87.4Recall@10DRL*
27.28842.89458.574.106Dec 17, 2021Apr 20, 2022Aug 22, 2022Dec 24, 2022Apr 27, 2023Aug 29, 2023Jan 1, 2024

Evaluation results

MethodLinks
DRL*Backbone=ViT-B/16, FLOPs (k=1000)=220.4k, DSL/Q-Norm trick=true2024.0187.456.279.9—74.5
EERCF*Backbone=ViT-B/16, FLOPs (k=1000)=0.8k, DSL/Q-Norm trick=true2024.0185.75577.8—72.8
DRLBackbone=ViT-B/16, FLOPs (k=1000)=220.4k2024.0185.448.976.3—70.2
X-CLIPBackbone=ViT-B/16, FLOPs (k=1000)=220.9k, reproduced=true2024.0184.548.976.8—70.1
X2-VLM (base)# Params=255M, mode=fine-tuning2022.1184.247.674.1——
X2-VLM (large)# Params=593M, mode=fine-tuning2022.1184.249.676.7——
EERCFBackbone=ViT-B/16, FLOPs (k=1000)=0.8k2024.0184.247.875.3—69.1
X-PoolBackbone=ViT-B/32, FLOPs (k=1000)=275.0k2024.018444.473.3—67.2
EERCFBackbone=ViT-B/32, FLOPs (k=1000)=16.0k2024.0183.944.774.2—67.6
OmniVL# Params=288M, mode=fine-tuning2022.1183.847.874.2——
DRLBackbone=ViT-B/32, FLOPs (k=1000)=220.4k2024.0183.345.373.9—67.5
TS2-NetBackbone=ViT-B/32, FLOPs (k=1000)=6.1k, reproduced=true2024.0182.743.671.1—65.8
CLIP4ClipBackbone=ViT-B/16, FLOPs (k=1000)=0.5k, reproduced=true2024.0182.445.473.4—67.1
CenterCLIPBackbone=ViT-B/32, FLOPs (k=1000)=1.5k2024.0181.742.971.4—65.3
X-CLIPBackbone=ViT-B/32, FLOPs (k=1000)=220.9k, reproduced=true2024.0181.245.972.8—66.7
CLIP4ClipBackbone=ViT-B/32, FLOPs (k=1000)=0.5k, reproduced=true2024.0180.541.470.6—64.2
CLIP-VIPBackbone=ViT-B/32, FLOPs (k=1000)=0.5k2024.0179.340.670.4—63.4
All-in-one# Params=110M, mode=fine-tuning2022.1177.137.968.1——
mPLUGEvaluation Protocol=Zero-Shot, # Pretrain data=14M, Finetuned on COCO=true2022.0575.444.366.4——
TeachTextBackbone=ViT-B/32, FLOPs (k=1000)=0.8k2024.017532.162.7—56.6
BLIPEvaluation Protocol=zero-shot2022.0174.743.365.62—
BLIPEvaluation Protocol=Zero-Shot, # Pretrain data=129M, Finetuned on COCO=true2022.0574.743.365.6——
VIOLETEvaluation Protocol=Fine-Tuning, # Pretrain data=Y180M, C3M, W2M2022.0573.434.563——
ALPROEvaluation Protocol=Finetuning, PT datasets=Web2M + CC3M (5.5M)2021.1273.233.960.73—
ALPROEvaluation Protocol=Fine-Tuning, # Pretrain data=C3M, W2M2022.0573.233.960.7——
FlorenceEvaluation Protocol=Zero-Shot, # Pretrain data=FLD900M2022.0572.637.663.8——
FiTEvaluation Protocol=Finetuning, PT datasets=Web2M + CC3M (5.5M)2021.1270.53159.53—
SupportSetEvaluation Protocol=Finetuning, PT datasets=HT (100M)2021.1269.330.158.53—
mPLUGEvaluation Protocol=Zero-Shot, # Pretrain data=14M2022.0568.238.159.2——
VideoClipEvaluation Protocol=Finetuning, PT datasets=HT (100M)2021.1266.830.955.4——
VideoCLIPEvaluation Protocol=finetuning2022.0166.830.955.4——
VideoCLIPEvaluation Protocol=Fine-Tuning, # Pretrain data=How100M2022.0566.830.955.4——
AVLNetEvaluation Protocol=Finetuning, PT datasets=HT (100M)2021.1266.627.155.64—
CLIPEvaluation Protocol=Zero-Shot, # Pretrain data=WIT400M2022.0560.72649.4——
ClipBERTEvaluation Protocol=Finetuning, PT datasets=COCO + VG (5.6M)2021.1259.92246.86—
ClipBERTEvaluation Protocol=finetuning2022.0159.92246.86—
VIOLETEvaluation Protocol=Zero-Shot, # Pretrain data=Y180M, W2M, C3M2022.0559.725.949.5——
HEROEvaluation Protocol=Finetuning, PT datasets=HT (100M)2021.1257.716.843.4——
ActBERTEvaluation Protocol=Finetuning, PT datasets=HT (100M)2021.1256.916.342.810—
PTP-VILTZero-shot=true2022.1256.327.952.57—
OA-TransZero-shot=true2022.1255.623.447.58—
ALPROEvaluation Protocol=Zero-shot, PT datasets=Web2M + CC3M (5.5M)2021.1255.424.144.78—
ALPROEvaluation Protocol=Zero-Shot, # Pretrain data=W2M, C3M2022.0555.424.144.7——
NoiseEst.Evaluation Protocol=Finetuning, PT datasets=HT (100M)2021.1253.617.441.68—
HT100MEvaluation Protocol=Finetuning, PT datasets=HT (100M)2021.1252.814.940.29—
FiTEvaluation Protocol=Zero-shot, PT datasets=Web2M + CC3M (5.5M)2021.1251.618.739.510—
FiTEvaluation Protocol=zero-shot2022.0151.618.739.510—
Frozen-in-timeZero-shot=true2022.1251.618.739.510—
JSFusionEvaluation Protocol=Finetuning, PT datasets=None2021.1243.210.231.213—
ActBERTEvaluation Protocol=Zero-shot, PT datasets=HT (100M)2021.1233.18.623.436—
ActBERTEvaluation Protocol=zero-shot2022.0133.18.623.436—
ActBERTZero-shot=true2022.1233.18.623.436—
MIL-NCEEvaluation Protocol=Zero-shot, PT datasets=HT (100M)2021.1232.49.92429.5—
MIL-NCEEvaluation Protocol=zero-shot2022.0132.49.92429.5—
MIL-NCEZero-shot=true2022.1232.49.92429.5—
MIL-NCEEvaluation Protocol=Zero-Shot, # Pretrain data=How100M2022.0532.49.924——
SupportSetEvaluation Protocol=Zero-shot, PT datasets=HT (100M)2021.1231.18.72331—
SupportSetEvaluation Protocol=zero-shot2022.0131.18.72331—
VideoCLIPEvaluation Protocol=Zero-shot, PT datasets=HT (100M)2021.123010.422.2——
VideoCLIPEvaluation Protocol=zero-shot2022.013010.422.2——
VideoCLIPEvaluation Protocol=Zero-Shot, # Pretrain data=How100M2022.053010.422.2——
VATTEvaluation Protocol=Zero-Shot, # Pretrain data=How100M, AudSet2022.0529.7————
HT100MEvaluation Protocol=Zero-shot, PT datasets=HT (100M)2021.1229.67.521.238—