List your company
ResearchFormerly WizWand

Scene Text Recognition on CUTE (test)

93.5AccuracyMATRN
49.164860.674972.18583.6951Mar 27, 2020Aug 23, 2020Jan 19, 2021Jun 17, 2021Nov 13, 2021Apr 11, 2022Sep 8, 2022
Updated 8mo ago

Evaluation results

MethodLinks
MATRN2021.1193.5
PREN2DYear=20212021.1191.7
MGP-STR VisionTraining Data=MJ and ST, Lexicon=None2022.0990.63
RobustScannerYear=20202021.1190.3
RobustScannerTraining Data=MJ and ST, Lexicon=None2022.0990.3
MGP-STR FuseTraining Data=MJ and ST, Lexicon=None2022.0990.28
JVSRYear=20212021.1189.7
ABINetYear=20212021.1189.2
ABINetTraining Data=MJ and ST, Lexicon=None2022.0989.2
ABINet (reproduced)Implementation=reproduced2021.1189
VisionLANYear=20212021.1188.5
VisionLANTraining Data=MJ and ST, Lexicon=None2022.0988.5
Liao et al. (SAM)Backbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0387.8
SRNBackbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0387.8
SATRNYear=20202021.1187.8
SRNYear=20202021.1187.8
SATRNTraining Data=MJ and ST, Lexicon=None2022.0987.8
SRNTraining Data=MJ and ST, Lexicon=None2022.0987.8
Yang et al. (SCRN)Backbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word and character-level, Lexicon=None2020.0387.5
GA-SPINYear=20212021.1187.5
MASTERTraining Data=MJ and ST, Lexicon=None2022.0987.5
Lyu et al. (Parallel)Backbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0386.8
SAFL2022.0185.4
Yang et al.Year=20202021.1185.4
SRN w/o GSRMBackbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0384.7
DANYear=20202021.1184.4
DANTraining Data=MJ and ST, Lexicon=None2022.0984.4
STAR-NetTrain Dataset=S90k+ST+TextOCR2021.0583.62
SE-ASTERYear=20202021.1183.6
SE-ASTERTraining Data=MJ and ST, Lexicon=None2022.0983.6
Zhan et al. (ESIR)Backbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0383.3
Li et al. (SAR)Backbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0383.3
SAR2022.0183.3
ESIR2022.0183.3
ESIRYear=20192021.1183.3
ESIRTraining Data=MJ and ST, Lexicon=None2022.0983.3
TextScannerTraining Data=MJ and ST, Lexicon=None2022.0983.3
Xie et al. (ACE)Backbone=VGG, Training Data=Synth90K, Annotations=word-level, Lexicon=None2020.0382.6
ViTSTRTraining Data=MJ and ST, Lexicon=None2022.0981.3
Liao et al.2022.0179.9
Shi et al. (ASTER)Backbone=ResNet, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0379.5
ASTER2022.0179.5
TPS-ResNet-BiLSTM-AttnTrain Dataset=S90k+ST+TextOCR2021.0579.44
RosettaTrain Dataset=S90k+ST+TextOCR2021.0577.35
Cheng et al. (AON)Backbone=self-designed, Training Data=Synth90K+SynthText, Annotations=word-level, Lexicon=None2020.0376.8
AON2022.0176.8
TPS-ResNet-BiLSTM-AttnPW=true, Train Dataset=S90k+ST2021.0574.22
Baek et al.2022.0174
TBRATraining Data=MJ and ST, Lexicon=None2022.0974
CRNNTrain Dataset=S90k+ST+TextOCR2021.0573.87
STAR-NetPW=true, Train Dataset=S90k+ST2021.0572.47
CombBestYear=20192021.1171
TPS-ResNet-BiLSTM-AttnTrain Dataset=TextOCR2021.0570.38
Yang et al.Backbone=VGG, Training Data=Synth90K+self-made, Annotations=word and character-level, Lexicon=None2020.0369.3
Yang et al.2022.0169.3
RosettaPW=true, Train Dataset=S90k+ST2021.0567.6
CRNNPW=true, Train Dataset=S90k+ST2021.0562.37
RARE2022.0159.2
TPS-ResNet-BiLSTM-AttnTrain Dataset=COCOText2021.0550.87