List your company

Scene Text Recognition on IC15 (test)

93.5Word AccuracyDTrOCR
82.5885.41588.2591.085May 12, 2021Jan 1, 2022Aug 23, 2022Apr 14, 2023Dec 4, 2023Jul 25, 2024Mar 17, 2025

Evaluation results

MethodLinks
DTrOCRTraining data=Real2023.0893.5
DTrOCRTraining data=Real2023.0893.2
CLIP4STR-L*Training data=REBU-Syn2023.1292.6
PARSeq-BTrain Data=REBU-Syn2023.1292.44
PARSeq-LTrain Data=REBU-Syn2023.1292.32
DTrOCRTraining data=Synth2023.0892.3
CLIP4STR-B*Training data=REBU-Syn2023.1292.27
PARSeq-SParameters=22M, Train Data=REBU-Syn2023.1292.27
CLIP4STR-LTraining data=Real2023.1291.66
PARSeq-HParameters=0.6B, Train Data=REBU-Syn2023.1291.66
CLIP4STR-BTraining data=Real2023.1291.39
PARSeq_ATrain data=R2022.0790.7
PARSeqTraining data=Real2023.0890.7
MAERec-BTraining data=Union14M-L2023.1290.7
ABINetTraining data=Real2023.1290.5
DTrOCRTraining data=Synth2023.0890.4
PARSeqATraining data=Real, Reproduced=true2023.1290.34
ABINetTrain data=R2022.0790.2
ABINetTraining data=Real2023.0890.2
MaskOCR(ViT-L)Training data=MJ+ST2023.1290.1
TRBATrain data=R2022.0789.8
TRBATraining data=Real2023.0889.8
PARSeq_NTrain data=R2022.0789.6
PARSeq_ATrain data=R2022.0789.6
PARSeqTraining data=Real2023.0889.6
ViTSTR-STraining data=Real2023.1289
DiG-ViT-BTraining data=Real2023.1288.9
TRBATrain data=R2022.0788.7
MaskOCRBackbone=ViT-L2021.0988.7
MaskOCR_LARGETraining data=Synth2023.0888.7
TRBATraining data=Real2023.0888.7
ABINetTrain data=R2022.0788.5
ABINetTraining data=Real2023.0888.5
VITSTR-STrain data=R2022.0788.4
PARSeq_NTrain data=R2022.0788.4
TrOCRBackbone=TrOCR-Large, Training Data=Synthetic + Benchmark2021.0988.1
MaskOCRBackbone=ViT-B2021.0987.3
MaskOCR_BASETraining data=Synth2023.0887.3
MaskOCR(ViT-B)Training data=MJ+ST2023.1287.3
VITSTR-STrain data=R2022.0787.1
DiG-ViT-BTraining data=MJ+ST2023.1287.1
TrOCRBackbone=TrOCR-Base, Training Data=Synthetic + Benchmark2021.0986.9
SVTR_LARGETraining data=Synth2023.0886.6
PARSeq_ATrain data=S2022.0786.5
PARSeq2021.0986.5
PARSeqTraining Data=90K+ST, Params=23.8M, Time=11.8ms2023.0586.5
ABINet-LV+TPS++Training Data=90K+ST, Params=37.2M, Time=41.5ms2023.0586.5
PARSeqTraining data=Synth2023.0886.5
STN-CSTRTrain data=S2022.0786.1
STN-CSTR2021.0986.1
STAR-NetTrain Dataset=S90k+ST+TextOCR2021.0586.09
ABINetTrain data=S32022.0786
ABINet2021.0986
ABINet-LVTraining Data=90K+ST, Params=36.7M, Time=37.2ms2023.0586
ABINetTraining data=Synth2023.0886
DiffusionSTRTraining data=Synth2023.0886
ABINetYear=2021, Training Data=90K+ST2021.1186
CDistNetTraining Data=90K+ST2021.1186
ABINetTraining data=MJ+ST2023.1286
NRTR+TPS++Training Data=90K+ST, Params=35.5M, Time=218ms2023.0585.7
ABINet + SeTaPruned %=28.1%2025.0385.6
ABINetPruned %=0%2025.0385.4
ABINet + InfoBatchPruned %=26.6%2025.0385.4
ABINet + SeTaPruned %=40.4%2025.0385.4
SVTR_BASETraining data=Synth2023.0885.2
CDistNet w/o TPSTraining Data=90K+ST2021.1185.2
TPS-ResNet-BiLSTM-AttnTrain Dataset=S90k+ST+TextOCR2021.0585.15
PARSeq_NTrain data=S2022.0785.1
ABINet + SeTaPruned %=71.0%2025.0384.9
CVAE-FeedTrain data=S2022.0784.6
CVAE-Feed.2021.0984.6
GTRTraining Data=90K+ST, Params=42.1M, Time=18.8ms2023.0584.6
CVAE-FeedTraining data=Synth2023.0884.6
S-GTRYear=2022, Training Data=90K+ST2021.1184.6
TRBATrain data=S2022.0784.3
ABINet + InfoBatchPruned %=38.1%2025.0384.2
TrOCRBackbone=TrOCR-Large, Training Data=Synthetic + Benchmark2021.0984.1
TrOCRLargeTraining data=MJ+ST+B2023.1284.1
Bhunia et al.Train data=S2022.0784
TrOCRBackbone=TrOCR-Large, Training Data=Synthetic2021.0984
Bhunia2021.0984
TrOCR_LARGETraining data=Synth2023.0884
JVSRTraining data=Synth2023.0884
JVSRYear=2021, Training Data=90K+ST2021.1184
ABINet + InfoBatchPruned %=50.3%2025.0384
VisionLANTrain data=S2022.0783.7
VisionLAN2021.0983.7
VisionLANTraining Data=90K+ST, Params=32.8M, Time=28.0ms2023.0583.7
VisionLANTraining data=Synth2023.0883.7
VisionLANYear=2021, Training Data=90K+ST2021.1183.7
VisionLANTraining data=MJ+ST2023.1283.7
TextScannerTrain data=S*2022.0783.5
TextScanner2021.0983.5
TextScannerTraining data=Synth2023.0883.5
TextScannerTraining data=MJ+ST2023.1283.5
PETRTraining data=MJ+ST2023.1283.3
ABINetTrain data=S2022.0783.1
PREN2DTrain data=S2022.0783
PREN2D2021.0983
Pren2DTraining Data=90K+ST+Real, Time=67.4ms2023.0583