List your company
ResearchFormerly WizWand

Scene Text Recognition on CUTE80 (test)

0.9965AccuracyCLIP4STR-B*
0.404220.5579850.711750.865515Mar 12, 2016Sep 11, 2017Mar 14, 2019Sep 13, 2020Mar 15, 2022Sep 15, 2023Mar 17, 2025
Updated 8mo ago

Evaluation results

MethodLinks
CLIP4STR-B*Training data=REBU-Syn2023.120.9965
CLIP4STR-L*Training data=REBU-Syn2023.120.9965
CLIP4STR-BTraining data=Real2023.120.9896
CLIP4STR-LTraining data=Real2023.120.9896
MAERec-BTraining data=Union14M-L2023.120.986
PARSeqATraining data=Real, Reproduced=true2023.120.9722
ABINetTraining data=Real2023.120.972
DiG-ViT-BTraining data=Real2023.120.965
ViTSTR-STraining data=Real2023.120.962
MaskOCR(ViT-L)Training data=MJ+ST2023.120.958
TrOCRLargeTraining data=MJ+ST+B2023.120.951
MATRNTraining data=MJ+ST2023.120.935
RobustScannerLexicon-free=true, Character-level annotations=false2021.060.924
NRTR+TPS++Training Data=90K+ST, Params=35.5M, Time=218ms2023.050.924
GTRTraining Data=90K+ST, Params=42.1M, Time=18.8ms2023.050.923
PARSeqTraining Data=90K+ST, Params=23.8M, Time=11.8ms2023.050.922
PARSeqATraining data=MJ+ST2023.120.922
RCEEDLexicon-free=true, Character-level annotations=false2021.060.917
Pren2DTraining Data=90K+ST+Real, Time=67.4ms2023.050.917
TextScannerTraining data=MJ+ST2023.120.916
DiG-ViT-BTraining data=MJ+ST2023.120.913
RobustScannerTraining Data=90K+ST2023.050.903
PETRTraining data=MJ+ST2023.120.899
SARLexicon=None, Training Level=word-level2018.110.896
SARLexicon-free=true, Character-level annotations=false2021.060.896
ABINet-LV+TPS++Training Data=90K+ST, Params=37.2M, Time=41.5ms2023.050.896
ABINet + InfoBatchPruned %=26.6%2025.030.894
ABINet-LVTraining Data=90K+ST, Params=36.7M, Time=37.2ms2023.050.892
TRBATraining data=MJ+ST2023.120.892
ABINetTraining data=MJ+ST2023.120.892
MaskOCR(ViT-B)Training data=MJ+ST2023.120.892
ABINetPruned %=0%2025.030.892
ABINet + SeTaPruned %=28.1%2025.030.892
Mask TextSpotterLexicon-free=true, Character-level annotations=true2021.060.885
VisionLANTraining Data=90K+ST, Params=32.8M, Time=28.0ms2023.050.885
VisionLANTraining data=MJ+ST2023.120.885
ABINet + InfoBatchPruned %=50.3%2025.030.885
ABINet + SeTaPruned %=40.4%2025.030.885
ABINet + InfoBatchPruned %=38.1%2025.030.884
SGBANetTraining Data=90K+ST2023.050.882
SATRNFeature map=2D, Training data=MJ+ST, Dictionary matching=False2019.100.878
SRNTraining Data=90K+ST, Params=49.3M, Time=26.9ms2023.050.878
SRNTraining data=MJ+ST2023.120.878
MASTERLexicon-free=true, Character-level annotations=false2021.060.875
SAR 3 x 20Lexicon-free=true, Character-level annotations=false2021.060.875
SPINTraining Data=90K+ST2023.050.875
ABINet + SeTaPruned %=71.0%2025.030.875
Yang, L et al.Lexicon-free=true, Character-level annotations=false2021.060.854
CRNN+TPS++Training Data=90K+ST, Params=16.2M, Time=16.6ms2023.050.854
SCATTERLexicon-free=true, Character-level annotations=false2021.060.851
SE-ASTERTraining data=MJ+ST2023.120.836
ESIRFeature map=STN-1D, Training data=MJ+ST, Dictionary matching=False2019.100.833
SARFeature map=2D, Training data=MJ+ST, Dictionary matching=False2019.100.833
ESIRLexicon-free=true, Character-level annotations=false2021.060.833
SARTraining Data=90K+ST, Params=57.5M, Time=120ms2023.050.833
ESIRTraining Data=90K+ST2023.050.833
STANTraining Data=90K+ST2023.050.833
ViTSTR-BTraining data=MJ+ST2023.120.813
NRTRTraining Data=90K+ST, Params=31.7M, Time=212ms2023.050.809
CA-FCNFeature map=2D, Training data=ST+C, Dictionary matching=False2019.100.799
Shi et al. 2018Lexicon=None, Training Level=word-level2018.110.795
ASTERFeature map=STN-1D, Training data=MJ+ST, Dictionary matching=False2019.100.795
ASTERLexicon-free=true, Character-level annotations=false2021.060.795
ASTERTraining Data=90K+ST, Params=22M, Time=73.1ms2023.050.795
Liu, Chen, and Wong 2018Lexicon=None, Training Level=word-level and character-level2018.110.789
MORANLexicon-free=true, Character-level annotations=false2021.060.774
MORANTraining Data=90K+ST, Params=28.5M, Time=24.4ms2023.050.774
AON (Ours)Lexicon=None2017.110.768
Cheng et al. 2018Lexicon=None, Training Level=word-level2018.110.768
AONFeature map=2D, Training data=MJ+ST, Dictionary matching=False2019.100.768
Yang et al. 2017Lexicon=None, Training Level=word-level and character-level2018.110.758
Naive_baseLexicon=None2017.110.754
Comb.BestFeature map=STN-1D, Training data=MJ+ST, Dictionary matching=False2019.100.74
STN_baseLexicon=None2017.110.737
Yang et al.Lexicon=None2017.110.693
ATRFeature map=2D, Training data=PRI+C, Dictionary matching=False2019.100.693
Cheng et al.Lexicon=None2017.110.639
Cheng et al. 2017Lexicon=None, Training Level=word-level and character-level2018.110.639
RARELexicon=None2016.030.592
Shi et al. [32]Lexicon=None2017.110.592
Shi et al. 2016Lexicon=None, Training Level=word-level2018.110.592
RAREFeature map=1D, Training data=MJ, Dictionary matching=False2019.100.592
Shi et al.Lexicon=None2016.030.549
Shi et al. [31]Lexicon=None2017.110.549
Shi, Bai, and Yao 2017Lexicon=None, Training Level=word-level2018.110.549
Jaderberg et al.Lexicon=None2016.030.427
Jaderberg et al. 2015aLexicon=None, Training Level=word-level2018.110.427