List your company

Text-to-Image Retrieval on MS-COCO 1k (test)

64.7R@1CRCL
11.76425.50739.2552.993Jan 13, 2021Jul 1, 2021Dec 17, 2021Jun 5, 2022Nov 21, 2022May 9, 2023Oct 26, 2023

Evaluation results

MethodLinks
CRCLNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1064.7—90.695.9525.6
MSCNNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1064.3—90.495.8524.6
BiCroNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1063.7—90.395.7523.2
VSRN + AOQD=2048 x 22021.0163.545.690.595.8—
VSRN + AOQ2021.0163.545.6———
CRCLNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1063.3—90.395.7521.5
VSRND=20482021.0162.842.489.795.1—
VSRN2021.0162.842.4———
BiCroNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1061.8—89.294.9517.1
DECLNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1061.6—89.195.2517.3
NCRNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1061.5—89.395.1516.6
DECLNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1061.5—89.295.3517.1
CRCLNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1060.8—8995.1514.2
NCRNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1060.7—88.595513.9
MSCNNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1060.3—88.594.4510.9
DECLNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1059.6—87.994.5508.2
MSCNNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1059.2—87.192.8506.5
BiCroNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1058.3—87.293.9505.5
NCRNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1057.9—8794.1503.3
CRCLNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1057.5—86.893.7501.8
VSE∞Noise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1057.4—86.592.8500.5
PVSEK=2, D=1024 x 22021.0155.243.786.593.7—
PVSEK=22021.0155.243.6———
PCME (mu only)D=10242021.0154.645.986.393.8—
PCMED=1024 x 22021.0154.64686.393.8—
PCMEvariant=mu only2021.0154.645.9———
PCMEvariant=stochastic2021.0154.646———
PVSEK=4, D=1024 x 42021.0154.142.785.592.9—
PVSE + SHMK=1, D=1024 x 12021.0153.642.485.593—
PVSEK=1, D=10242021.0153.541.985.192.7—
PVSEK=12021.0153.541.8———
PVSE + SHMK=2, D=1024 x 22021.0153.139.485.493—
MSCNNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1052.7—8390.9481.2
VSE++D=10242021.0152—84.392—
VSE++2021.0152————
DECLNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1052—8391.3480.1
SGRNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1042.5—77.788.2434
SAFNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1038.2—7485.5406.5
SGRNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1037—72.885.1406.4
SGRNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1033.8—68.681.7381.3
SAFNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1033.7—69.482.5383
VSE∞Noise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1031.4—63.875399.9
SAFNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1031.1—65.680.5348.7
VSE∞Noise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1026—60.176.3339.7
SGRNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1025.3—58.272.6319.1
SAFNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1024.7—57.173310.6
BiCroNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1022.6—5367.2340.1
SCANNoise Rate=20%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1020.6—55.673.5322.6
VSE∞Noise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1019.2—50.568288.8
SCANNoise Rate=40%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1018.9—51.169.9314.4
SCANNoise Rate=60%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1016.8—47.866.4293.4
NCRNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1015.1—38.149.8244.8
SCANNoise Rate=80%, Visual Feature=BUTD, Backbone=Bi-GRU, Evaluation=5-fold average2023.1013.8—41.158.6255.1