List your company
ResearchFormerly WizWand

Scene Text Recognition on IC13 (test)

99.42Word AccuracyCLIP4STR-L*
94.927296.093697.2698.4264Mar 27, 2020Nov 10, 2020Jun 27, 2021Feb 11, 2022Sep 27, 2022May 14, 2023Dec 29, 2023
Updated 8mo ago

Evaluation results

MethodLinks
CLIP4STR-L*Training data=REBU-Syn2023.1299.42
DTrOCRTraining data=Real2023.0899.4
CLIP4STR-B*Training data=REBU-Syn2023.1299.29
DTrOCRTraining data=Real2023.0899.1
DTrOCRTraining data=Synth2023.0898.8
CLIP4STR-LTraining data=Real2023.1298.48
PARSeq_ATrain data=R2022.0798.4
PARSeqTraining data=Real2023.0898.4
CLIP4STR-BTraining data=Real2023.1298.36
PARSeq_ATrain data=R2022.0798.3
TrOCR_LARGETraining data=Synth2023.0898.3
PARSeqTraining data=Real2023.0898.3
MaskOCR(ViT-L)Training data=MJ+ST2023.1298.2
PARSeq_NTrain data=R2022.0798.1
MaskOCR_BASETraining data=Synth2023.0898.1
MaskOCR(ViT-B)Training data=MJ+ST2023.1298.1
MAERec-BTraining data=Union14M-L2023.1298.1
ABINetTrain data=R2022.0798
PARSeq_NTrain data=R2022.0798
ABINetTraining data=Real2023.0898
ABINetTraining data=Real2023.0898
ABINetTraining data=Real2023.1298
ABINetTrain data=R2022.0797.8
ABINet-LV+TPS++Training Data=90K+ST, Params=37.2M, Time=41.5ms2023.0597.8
MaskOCR_LARGETraining data=Synth2023.0897.8
DTrOCRTraining data=Synth2023.0897.8
ViTSTR-STraining data=Real2023.1297.8
VITSTR-STrain data=R2022.0797.7
VITSTR-STrain data=R2022.0797.6
TRBATrain data=R2022.0797.6
TRBATrain data=R2022.0797.6
TRBATraining data=Real2023.0897.6
TRBATraining data=Real2023.0897.6
DiG-ViT-BTraining data=Real2023.1297.6
ABINetTrain data=S32022.0797.4
TrOCRBackbone=TrOCR-Base, Training Data=Synthetic + Benchmark2021.0997.4
ABINet-LVTraining Data=90K+ST, Params=36.7M, Time=37.2ms2023.0597.4
ABINetTraining data=Synth2023.0897.4
ABINetYear=2021, Training Data=90K+ST2021.1197.4
CDistNet w/o TPSTraining Data=90K+ST2021.1197.4
CDistNetTraining Data=90K+ST2021.1197.4
ABINetTraining data=MJ+ST2023.1297.4
PARSeqATraining data=Real, Reproduced=true2023.1297.32
TrOCRBackbone=TrOCR-Large, Training Data=Synthetic + Benchmark2021.0997.3
TrOCR_BASETraining data=Synth2023.0897.3
TrOCRLargeTraining data=MJ+ST+B2023.1297.3
SVTR_LARGETraining data=Synth2023.0897.2
ABINetTrain data=S2022.0797.1
DiffusionSTRTraining data=Synth2023.0897.1
SVTR_BASETraining data=Synth2023.0897.1
PARSeq_ATrain data=S2022.0797
TrOCRBackbone=TrOCR-Large, Training Data=Synthetic2021.0997
PARSeqTraining Data=90K+ST, Params=23.8M, Time=11.8ms2023.0597
PARSeqTraining data=Synth2023.0897
TrOCR_LARGETraining data=Synth2023.0897
PETRTraining data=MJ+ST2023.1297
DiG-ViT-BTraining data=MJ+ST2023.1296.9
GTRTraining Data=90K+ST, Params=42.1M, Time=18.8ms2023.0596.8
S-GTRYear=2022, Training Data=90K+ST2021.1196.8
NRTR+TPS++Training Data=90K+ST, Params=35.5M, Time=218ms2023.0596.6
PREN2DTrain data=S2022.0796.4
Pren2DTraining Data=90K+ST+Real, Time=67.4ms2023.0596.4
PREN2DTraining data=Synth2023.0896.4
DiffusionSTRTraining data=Synth2023.0896.4
Pren2DYear=2021, Training Data=90K+ST+Real2021.1196.4
STN-CSTRTrain data=S2022.0796.3
TRBATrain data=S2022.0796.3
PARSeq_NTrain data=S2022.0796.3
TrOCRBackbone=TrOCR-Base, Training Data=Synthetic2021.0996.3
TrOCR_BASETraining data=Synth2023.0896.3
PARSeq_ATrain data=S2022.0796.2
PARSeq2021.0996.2
PARSeqTraining data=Synth2023.0896.2
PARSeqATraining data=MJ+ST2023.1296.2
NRTRTraining Data=90K+ST, Params=31.7M, Time=212ms2023.0595.8
NRTRYear=2019, Training Data=90K+ST2021.1195.8
MATRNTraining data=MJ+ST2023.1295.8
VisionLANTrain data=S2022.0795.7
CVAE-FeedTrain data=S2022.0795.7
CVAE-Feed.2021.0995.7
VisionLANTraining Data=90K+ST, Params=32.8M, Time=28.0ms2023.0595.7
VisionLANTraining data=Synth2023.0895.7
CVAE-FeedTraining data=Synth2023.0895.7
VisionLANYear=2021, Training Data=90K+ST2021.1195.7
VisionLANTraining data=MJ+ST2023.1295.7
SRNBackbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0395.5
SRNTrain data=S2022.0795.5
Bhunia et al.Train data=S2022.0795.5
PARSeq_NTrain data=S2022.0795.5
Bhunia2021.0995.5
SRNTraining Data=90K+ST, Params=49.3M, Time=26.9ms2023.0595.5
SRNTraining data=Synth2023.0895.5
JVSRTraining data=Synth2023.0895.5
SRNYear=2020, Training Data=90K+ST2021.1195.5
JVSRYear=2021, Training Data=90K+ST2021.1195.5
SRNTraining data=MJ+ST2023.1295.5
STAR-NetTrain Dataset=S90k+ST+TextOCR2021.0595.33
Liao et al. (SAM)Backbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0395.3
MaskTextSpotterYear=2019, Training Data=90K+ST2021.1195.3
VITSTR-STrain data=S2022.0795.1