List your company

Text-to-Video Retrieval on MSR-VTT (9K)

52R@1TEFAL + DSL + QB-Norm
1.5614.65527.7540.845Jan 20, 2022Apr 21, 2022Jul 22, 2022Oct 22, 2022Jan 21, 2023Apr 23, 2023Jul 24, 2023

Evaluation results

MethodLinks
TEFAL + DSL + QB-NormBackbone=ViT-B/16, Modality (Audio)=true, Post-processing (DSL)=true, Post-processing (QB-Norm)=true2023.075276.686.1111.4
TEFAL + DSLBackbone=ViT-B/16, Modality (Audio)=true, Post-processing (DSL)=true2023.0750.17785.4110.5
TEFALBackbone=ViT-B/16, Modality (Audio)=true2023.0749.976.284.4211.4
TEFALBackbone=ViT-B/32, Modality (Audio)=true2023.0749.475.983.9212
CenterCLIPMeM. GB=17.6, Speed ms=86.5, Backbone=ViT-B/16, Clustering=k-medoids++, Budget=B6 - 4, 1602022.0548.473.882213.8
OmniVLBackbone=ViT-B/16, Additional training pairs=true2023.0747.874.283.8——
VoP^F+CParams (M)=14.1, Backbone=ViT-B/162022.1147.772.482.2212
CAMOE + DSLBackbone=ViT-B/32, Post-processing (DSL)=true2023.0747.374.683.8211.9
VoP^F+PParams (M)=0.4, Backbone=ViT-B/162022.1147.172.481.8212.9
TS2-NetBackbone=ViT-B/322023.074774.583.8213
X-PoolBackbone=ViT-B/322023.0746.972.882.2214.3
VoP^FParams (M)=0.1, Backbone=ViT-B/162022.1146.57381.5212.4
X-CLIPBackbone=ViT-B/322023.0746.17383.1213.2
CLIP4clip (meanP)MeM. GB=25.7, Speed ms=59.6, Backbone=ViT-B/16, variant=baseline2022.0545.671.280.9215.2
FullParams (M)=118.1, Backbone=ViT-B/162022.1144.972.281.7213.5
ECLIPSE_meanPBackbone=ViT-B/32, Modality (Audio)=true2023.0744.971.381.6215
BridgeFormerBackbone=ViT-B/32, Additional training pairs=true2023.0744.971.980.3215.3
VoP^CParams (M)=14.3, Backbone=ViT-B/162022.1144.671.880.2214.6
CLIP4clip (seqTransf)2022.0544.571.481.6215.3
CenterCLIPMeM. GB=14.2, Speed ms=43.6, Backbone=ViT-B/32, Clustering=spectral, Budget=B6 - 3, 492022.0544.271.682.1215.1
CenterCLIPBackbone=ViT-B/322023.0744.271.682.1215.1
CenterCLIPMeM. GB=14.2, Speed ms=22.9, Backbone=ViT-B/32, Clustering=k-medoids++, Budget=B6 - 3, 492022.054470.781.4215.7
VoP^PParams (M)=0.5, Backbone=ViT-B/162022.1143.97080.9212.9
CenterCLIPMeM. GB=15, Speed ms=22.9, Backbone=ViT-B/32, Clustering=k-medoids++, Budget=B6 - 4, 492022.0543.671.481.2215.3
CenterCLIPMeM. GB=14.9, Speed ms=40.8, Backbone=ViT-B/32, Clustering=spectral, Budget=B6 - 4, 492022.0543.671.780.6215.4
VoPParams (M)=0.1, Backbone=ViT-B/162022.1143.469.180.5214.2
CLIP4clip (meanP)MeM. GB=20.8, Speed ms=24.42022.0543.170.480.8216.2
CLIP4Clip_meanPBackbone=ViT-B/322023.0743.170.480.8215.3
CLIP4clip (meanP)MeM. GB=20.8, Speed ms=24.4, Backbone=ViT-B/32, variant=baseline2022.054370.780.6216.2
PartialParams (M)=7.7, Backbone=ViT-B/162022.114369.378.5215.8
BiasParams (M)=0.1, Backbone=ViT-B/162022.1142.268.578.2213.9
Adapter ATTNParams (M)=2.0, Backbone=ViT-B/162022.1141.766.476.6215.1
Adapter FFNParams (M)=2.0, Backbone=ViT-B/162022.1141.466.577215
ProjParams (M)=0.7, Backbone=ViT-B/162022.1139.165.775.6217.6
MV-GPTWith ASR=true2022.0137.365.575.12—
Ju et al.Params (M)=4.8, Backbone=ViT-B/162022.1136.764.676.82—
Bi-NCEWith ASR=true2022.0133.761.6733—
MV-GPTWith ASR=false2022.0133.661.273.63—
Bi-NCEWith ASR=false2022.0131.65970.23—
CLIP zero-shot2022.0531.253.764.24—
HiTWith ASR=false2022.0130.760.973.2——
SSBWith ASR=false2022.0130.158.569.33—
TT-CE+2022.0529.661.674.23—
T2VLAD2022.0529.55970.14—
AVLnet2022.0527.155.666.64—
AVLnetWith ASR=false2022.0127.155.666.64—
MMT2022.0526.657.169.6424
MMTWith ASR=false2022.0126.657.169.64—
UniVLWith ASR=false2022.0121.249.663.16—
CE2022.0520.948.862.46—
HowTo100M2022.0514.940.252.89—
JSFusion2022.0510.231.243.213—
ActBERT2022.058.623.433.136—
No PTWith ASR=true2022.015.613.318.492—
No PTWith ASR=false2022.013.5812.1114—