Scene Text Detection on TotalText (test)
90.5RecallHierText
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| HierText2025.04 | 90.5 | 85.5 | — | — | 87.9 | |
| DAT-SEG2025.04 | 89.2 | 95 | — | — | 92 | |
| DAT-DET2025.04 | 88.2 | 94 | — | — | 91 | |
| ESTextSpotter2025.04 | 88.1 | 92 | — | — | 90 | |
| MixNetPretrain=true, FSNet=true, Center line=true2023.08 | 88 | 93 | — | 15.2 | 90.5 | |
| MixNetPretrain=true, FSNet=true, Center line=false2023.08 | 87 | 93.1 | — | 15.7 | 89.9 | |
| PGNetBackbone=ResNet-50-FPN, FPS=35.52022.04 | 86.8 | 85.5 | 86.1 | — | — | |
| DPText-DETRBackbone=Res502022.07 | 86.4 | 91.8 | 89 | — | — | |
| DPText-DETR2025.04 | 86.4 | 91.8 | — | — | 89 | |
| Feature Sampling and Groupingrepresentation=Bezier2022.03 | 85.7 | 90.7 | 88.1 | — | — | |
| FSGBackbone=Res502022.07 | 85.7 | 90.7 | 88.1 | — | — | |
| UHT V16Backbone=VGG-16, Testing Scale=Multi-scale2019.11 | 85.6 | 85 | 85.3 | — | — | |
| SIR2025.04 | 85.6 | 90.9 | — | — | 88.2 | |
| CRAFTSMulti-scale inference=false2022.03 | 85.4 | 89.5 | 87.4 | — | — | |
| TextBPN++Backbone=Res502022.07 | 85.3 | 91.8 | 88.5 | — | — | |
| TextBPN2022.03 | 85.2 | 90.7 | 87.9 | — | — | |
| CharNet H-88Venue=ICCV-2019, Testing Scale=Multi-scale2019.11 | 85 | 88 | 86.5 | — | — | |
| CharNetMulti-scale inference=true2022.03 | 85 | 88 | 86.5 | — | — | |
| Qin et al.Multi-scale inference=false2022.03 | 85 | 87.8 | 86.4 | — | — | |
| CharNetBackbone=ResNet-50-Hourglass57, FPS=1.22022.04 | 85 | 87.3 | 86.1 | — | — | |
| Boundary TextSpotterBackbone=ResNet-50-FPN2022.04 | 85 | 88.9 | 87 | — | — | |
| DRRG2022.03 | 84.9 | 86.5 | 85.7 | — | — | |
| UHT R50Backbone=ResNet-50, Testing Scale=Multi-scale2019.11 | 84.2 | 85.1 | 84.8 | — | — | |
| ABCNet v2Multi-scale inference=true2022.03 | 84.1 | 90.2 | 87 | — | — | |
| ABCNet v2Backbone=ResNet-50-FPN, FPS=102022.04 | 84.1 | 90.2 | 87 | — | — | |
| ABCNet-v2Backbone=Res502022.07 | 84.1 | 90.2 | 87 | — | — | |
| MixNetPretrain=false, FSNet=true, Center line=false2023.08 | 84.1 | 92 | — | 15.2 | 87.8 | |
| DB + VLPT-STDPre-training=Ours (VLPT-STD)2022.04 | 84 | 88.7 | 86.3 | — | — | |
| MixNetPretrain=true, FSNet=false, Center line=false2023.08 | 84 | 91.8 | — | 21.7 | 87.7 | |
| ContourNetExternal Data=false, Testing Scale=single-scale, FPS=3.8, Evaluation Protocol=DetEval2020.04 | 83.9 | 86.9 | 85.4 | — | — | |
| ContourNet2022.03 | 83.9 | 86.9 | 85.4 | — | — | |
| TESTR-BezierBackbone=ResNet-50, FPS=5.52022.04 | 83.7 | 92.8 | 88 | — | — | |
| I3CLBackbone=Res502022.07 | 83.7 | 89.2 | 86.3 | — | — | |
| MixNetPretrain=false, FSNet=false, Center line=true2023.08 | 83.7 | 91.1 | — | 15.7 | 87.3 | |
| UnconstrainedBackbone=ResNet-50-MSF2022.04 | 83.4 | 83.3 | 83.3 | — | — | |
| DB + STKMPre-training=STKM2022.04 | 83.4 | 87.7 | 85.5 | — | — | |
| TextFuseNetBackbone=Res502022.07 | 83.2 | 87.5 | 85.3 | — | — | |
| UHT V16Backbone=VGG-16, Testing Scale=Single-scale2019.11 | 82.6 | 88.8 | 85.6 | — | — | |
| DocSAM2025.04 | 82.6 | 72.1 | — | — | 77 | |
| DB-ResNet-50Backbone=ResNet-50, Input Height=8002019.11 | 82.5 | 87.1 | 84.7 | 32 | — | |
| DB2022.03 | 82.5 | 87.1 | 84.7 | — | — | |
| DB + STPre-training=SynthText (ST)2022.04 | 82.5 | 87.1 | 84.7 | — | — | |
| DBBackbone=Res50-DCN2022.07 | 82.5 | 87.1 | 84.7 | — | — | |
| PCRBackbone=DLA342022.07 | 82 | 88.5 | 85.2 | — | — | |
| UHT R50Backbone=ResNet-50, Testing Scale=Single-scale2019.11 | 81.8 | 88.2 | 84.9 | — | — | |
| Text PerceptronBackbone=ResNet-50-FPN2022.04 | 81.8 | 88.8 | 85.2 | — | — | |
| CharNet H-88Venue=ICCV-2019, Testing Scale=Single-scale2019.11 | 81.7 | 89.9 | 85.6 | — | — | |
| DEERMulti-scale inference=false2022.03 | 81.44 | 90.44 | 85.71 | — | — | |
| TESTR-PolygonBackbone=ResNet-50, FPS=5.32022.04 | 81.4 | 93.4 | 86.9 | — | — | |
| TESTR-polygonBackbone=Res502022.07 | 81.4 | 93.4 | 86.9 | — | — | |
| PSENet_v2Venue=ICCV-2019, Testing Scale=Single-scale2019.11 | 81 | 89.3 | 85 | — | — | |
| PAN++Multi-scale inference=false2022.03 | 81 | 89.9 | 85.3 | — | — | |
| PAN2022.03 | 81 | 89.3 | 85 | — | — | |
| PANBackbone=Res-182022.07 | 81 | 89.3 | 85 | — | — | |
| ICGVenue=PR-2019, Testing Scale=Single-scale2019.11 | 80.9 | 82.9 | 81.5 | — | — | |
| MixNetPretrain=false, FSNet=false, Center line=false2023.08 | 80.5 | 91.4 | — | 21.7 | 85.7 | |
| TextField2019.11 | 79.9 | 81.2 | 80.6 | — | — | |
| CRAFT2019.11 | 79.9 | 87.6 | 83.6 | — | — | |
| TextFieldExternal Data=true, Testing Scale=single-scale, FPS=6, Evaluation Protocol=DetEval2020.04 | 79.9 | 81.2 | 80.6 | — | — | |
| CRAFTExternal Data=true, Testing Scale=single-scale, Evaluation Protocol=DetEval2020.04 | 79.9 | 87.6 | 83.6 | — | — | |
| TextFieldVenue=TIP-2019, Testing Scale=Single-scale2019.11 | 79.9 | 81.2 | 80.6 | — | — | |
| CRAFT+Venue=CVPR-2019, Testing Scale=Single-scale2019.11 | 79.9 | 87.6 | 83.6 | — | — | |
| TextField2022.03 | 79.9 | 81.2 | 80.6 | — | — | |
| CRAFT2022.03 | 79.9 | 87.6 | 83.6 | — | — | |
| CRAFTBackbone=VGG162022.07 | 79.9 | 87.6 | 83.6 | — | — | |
| CSE+Venue=CVPR-2019, Testing Scale=Single-scale2019.11 | 79.7 | 81.4 | 80.2 | — | — | |
| PANExternal Data=false, Testing Scale=single-scale, FPS=39.6, Evaluation Protocol=DetEval2020.04 | 79.4 | 88 | 83.5 | — | — | |
| LOMOmulti-scale=true2019.11 | 79.3 | 87.6 | 83.3 | — | — | |
| LOMOExternal Data=true, Testing Scale=multi-scale, Evaluation Protocol=DetEval2020.04 | 79.3 | 87.6 | 83.3 | — | — | |
| LOMOVenue=CVPR-2019, Testing Scale=Multi-scale2019.11 | 79.3 | 87.6 | 83.3 | — | — | |
| CSE2019.11 | 79.1 | 81.4 | 80.2 | — | — | |
| PSE-1s2019.11 | 78 | 84 | 80.9 | 3.9 | — | |
| PSENetExternal Data=true, Testing Scale=single-scale, FPS=3.9, Evaluation Protocol=DetEval2020.04 | 78 | 84 | 80.9 | — | — | |
| FTSNVenue=ICPR-2018, Testing Scale=Single-scale2019.11 | 78 | 84.7 | 81.3 | — | — | |
| PSE-Net2022.03 | 78 | 84 | 80.9 | — | — | |
| PSENet-1s+Venue=CVPR-2019, Testing Scale=Single-scale2019.11 | 77.96 | 84.02 | 80.87 | — | — | |
| DB-ResNet-18Backbone=ResNet-18, Input Height=8002019.11 | 77.9 | 88.3 | 82.8 | 50 | — | |
| ATRR2019.11 | 76.2 | 80.9 | 78.5 | — | — | |
| Wang et al.External Data=false, Testing Scale=single-scale, Evaluation Protocol=DetEval2020.04 | 76.2 | 80.9 | 78.5 | — | — | |
| LOMOExternal Data=false, Testing Scale=single-scale, FPS=4.4, Evaluation Protocol=DetEval2020.04 | 75.7 | 88.6 | 81.6 | — | — | |
| LOMOVenue=CVPR-2019, Testing Scale=Single-scale2019.11 | 75.7 | 88.6 | 81.6 | — | — | |
| TextDragonMulti-scale inference=false2022.03 | 75.7 | 85.6 | 80.3 | — | — | |
| LOMO2022.03 | 75.7 | 88.6 | 81.6 | — | — | |
| Text DragonBackbone=VGG162022.04 | 75.7 | 85.6 | 80.3 | — | — | |
| MTS2019.11 | 75.6 | 82.5 | 78.6 | — | — | |
| PSENetExternal Data=false, Testing Scale=single-scale, FPS=3.9, Evaluation Protocol=DetEval2020.04 | 75.1 | 81.8 | 78.3 | — | — | |
| MSRExternal Data=true, Testing Scale=single-scale, FPS=4.3, Evaluation Protocol=DetEval2020.04 | 74.8 | 83.8 | 79 | — | — | |
| TextSnake2019.11 | 74.5 | 82.7 | 78.4 | — | — | |
| TextSnakeExternal Data=true, Testing Scale=single-scale, Evaluation Protocol=DetEval2020.04 | 74.5 | 82.7 | 78.4 | — | — | |
| TextSnakeVenue=ECCV-2018, Testing Scale=Single-scale2019.11 | 74.5 | 82.7 | 78.4 | — | — | |
| TextSnake2022.03 | 74.5 | 82.7 | 78.4 | — | — | |
| TextSnakeBackbone=VGG162022.07 | 74.5 | 82.7 | 78.4 | — | — | |
| TextDragonExternal Data=true, Testing Scale=single-scale, Evaluation Protocol=DetEval2020.04 | 74.2 | 84.5 | 79 | — | — | |
| Poly-FRCNN-3Venue=IJDAR-2019, Testing Scale=Single-scale2019.11 | 68 | 78 | 73 | — | — | |
| TextNetMulti-scale inference=false2022.03 | 59.45 | 68.21 | 63.53 | — | — | |
| Li et al.Multi-scale inference=true2022.03 | 59.38 | 63.25 | 61.25 | — | — | |
| Lyu et al.External Data=true, Testing Scale=single-scale, Evaluation Protocol=DetEval2020.04 | 55 | 69 | 61.3 | — | — | |
| Mask TextSpotterMulti-scale inference=false2022.03 | 55 | 69 | 61.3 | — | — | |
| Mask TextSpotterBackbone=ResNet-50-FPN, FPS=4.82022.04 | 55 | 69 | 61.3 | — | — | |
| TextboxesBackbone=ResNet-50-FPN, FPS=1.42022.04 | 45.5 | 62.1 | 52.5 | — | — |