List your company
ResearchFormerly WizWand

Scene Text Recognition on IIIT5K 3,000 samples (test)

99.6Word AccuracyDTrOCR
81.08885.89490.795.506Sep 21, 2021Feb 6, 2022Jun 24, 2022Nov 9, 2022Mar 27, 2023Aug 12, 2023Dec 29, 2023
Updated 8mo ago

Evaluation results

MethodLinks
DTrOCRTraining data=Real2023.0899.6
PARSeqTraining data=Real2023.0899.1
PARSeq-HParameters=0.6B, Train Data=REBU-Syn2023.1298.93
PARSeq-SParameters=22M, Train Data=REBU-Syn2023.1298.63
PARSeq-LTrain Data=REBU-Syn2023.1298.63
TRBATraining data=Real2023.0898.6
ABINetTraining data=Real2023.0898.6
PARSeq-BTrain Data=REBU-Syn2023.1298.53
DTrOCRTraining data=Synth2023.0898.4
DiffusionSTRTraining Datasets=MJ and ST2023.0697.3
DiffusionSTRTraining data=Synth2023.0897.3
PARSeq2021.0997
PARSeqTraining Datasets=MJ and ST2023.0697
PARSeqTraining data=Synth2023.0897
Diaz2021.0996.8
MaskOCRBackbone=ViT-L2021.0996.5
MaskOCR_LARGETraining data=Synth2023.0896.5
SVTR_LARGETraining data=Synth2023.0896.3
ABINet2021.0996.2
ABINetTraining Datasets=MJ and ST2023.0696.2
ABINetTraining data=Synth2023.0896.2
SVTR_BASETraining data=Synth2023.0896
VisionLAN2021.0995.8
MaskOCRBackbone=ViT-B2021.0995.8
VisionLANTraining data=Synth2023.0895.8
MaskOCR_BASETraining data=Synth2023.0895.8
TextScanner2021.0995.7
TextScannerTraining data=Synth2023.0895.7
PREN2D2021.0995.6
PREN2DTraining data=Synth2023.0895.6
RobustScanner2021.0995.4
Bhunia2021.0995.2
CVAE-Feed.2021.0995.2
JVSRTraining data=Synth2023.0895.2
CVAE-FeedTraining data=Synth2023.0895.2
RCEED2021.0994.9
SRN2021.0994.8
SRNTraining data=Synth2023.0894.8
AutoSTR2021.0994.7
AutoSTRTraining data=Synth2023.0894.7
CRNNTraining data=Real2023.0894.6
PlugNet2021.0994.4
PlugNetTraining data=Synth2023.0894.4
STN-CSTR2021.0994.2
TrOCRBackbone=TrOCR-Large, Training Data=Synthetic + Benchmark2021.0994.1
TrOCRBackbone=TrOCR-Base, Training Data=Synthetic + Benchmark2021.0993.4
TRBA2021.0992.1
TRBATraining Datasets=MJ and ST2023.0692.1
TRBATraining data=Synth2023.0892.1
TrOCRBackbone=TrOCR-Large, Training Data=Synthetic2021.0991
TrOCR_LARGETraining data=Synth2023.0891
TrOCRBackbone=TrOCR-Base, Training Data=Synthetic2021.0990.1
TrOCR_BASETraining data=Synth2023.0890.1
VITSTR-B2021.0988.4
ViTSTRTraining Datasets=MJ and ST2023.0688.4
ViTSTR_BASETraining data=Synth2023.0888.4
CRNN2021.0984.3
CRNNTraining Datasets=MJ and ST2023.0681.8
CRNNTraining data=Synth2023.0881.8