method: CPN (multi-scale)2024-05-30

Authors: Longhuang Wu, Shangxuan Tian, Youxin Wang, Pengfei Xiong

Email: wlonghuang@gmail.com

Description: We propose a Complementary Proposal Network (CPN) that seamlessly and parallelly integrates semantic and geometric information for superior performance. This Result is achieved with single Swin-L backbone and multi-scale testing policy. No model ensemble is used.

method: AntFin-Cascade Mask R-CNN2023-02-23

Authors: Yangkun Lin, Tao Xu

Affiliation: Ant Group

Description: Our detector is based on Cascade Mask R-CNN. We use ConvNeXt-B as backbone. SynthText800k and
VISD10k are used to pretrain, and then we finetune on ArT, ICDAR2019-MLT and part of LSVT with multi-scale training. Multi-scale testing is used to get the result.

method: I3CL2021-07-05

Authors: Jian Ye, Jing Zhang, Juhua Liu, Bo Du and Dacheng Tao

Affiliation: Wuhan University SigmaLab, JD Explore Academy

Email: leaf-yej@whu.edu.cn

Description: A arbitrary-shaped scene text detector based on Mask R-CNN. In this result, we use ResNeSt-101 as the backbone. Multi-scale training and testing are applied to get the final result. Our training datasets contain SynthText (pretrain), ArT, ICDAR2019-MLT, and part of LSVT.

Ranking Table

Description Paper Source Code
DateMethodRecallPrecisionHmean
2024-05-30CPN (multi-scale)88.35%83.51%85.86%
2023-02-23AntFin-Cascade Mask R-CNN83.36%87.08%85.18%
2021-07-05I3CL81.03%87.26%84.03%
2020-05-21DuXiaoman_OCR79.35%87.81%83.36%
2019-12-17Tencent TEG OCR81.16%85.64%83.34%
2019-11-04Sogou_OCR78.49%87.94%82.95%
2019-04-30MEGVII_Detection76.68%89.64%82.65%
2020-04-22Mask R-CNN78.55%86.43%82.30%
2022-04-19TextBPN++(ResNet-50 with DCN)77.05%84.48%80.59%
2019-05-01NJU-ImagineLab74.21%87.35%80.24%
2019-04-29ArtDet-v273.54%86.45%79.48%
2023-08-09SRFormer (ResNet50-#1seg)73.51%86.08%79.30%
2022-10-31TD-PPIoU (Long-Pretrain)74.21%85.06%79.27%
2023-07-08CPNText-DETR(resnet-50)75.59%83.06%79.15%
2024-01-18LRANet74.51%84.06%79.00%
2022-04-21I3CL(ViTAEv2-S)75.42%82.82%78.95%
2023-03-08TD-PPIoU76.96%81.00%78.93%
2019-04-26baseline_polygon75.38%82.51%78.79%
2020-10-01TextFuseNet (ResNeXt-101)72.77%85.42%78.59%
2019-04-30CUTeOCR71.56%86.57%78.36%
2022-07-11DPText-DETR (ResNet-50)73.70%82.97%78.06%
2023-11-29ESRNet72.61%82.94%77.44%
2019-04-29Sg_ptd70.41%85.98%77.42%
2019-04-28 Alibaba-PAI73.25%79.18%76.10%
2022-03-25TextBPN++(ResNet-50)71.07%81.14%75.77%
2021-03-26TextFuseNet (ResNet-50)69.42%82.59%75.44%
2019-04-30Fudan-Supremind Detection v371.61%79.26%75.24%
2019-04-29SRCB_Art70.30%80.41%75.02%
2019-04-30A scene text detection method based on maskrcnn66.25%85.69%74.72%
2019-04-30DMText_art66.15%85.09%74.43%
2021-04-28NN_Chinese_and_euro666.51%82.74%73.74%
2019-04-30TEXT_SNIPER71.45%76.17%73.74%
2023-05-15dp_pq_nn66.27%82.60%73.54%
2019-04-28CLTDR65.92%82.58%73.32%
2021-04-08AutoCV69.59%77.25%73.22%
2019-04-29CRAFT68.93%77.25%72.85%
2019-04-30MaskRCNN_Text67.28%79.06%72.69%
2023-05-15dp_nn67.30%78.92%72.65%
2019-04-30QAQ63.45%83.76%72.21%
2019-04-30MaskDet67.04%76.47%71.44%
2019-04-24fdu_ai61.61%82.11%70.40%
2019-04-30CCISTD60.72%81.16%69.47%
2019-04-30Mask RCNN73.20%65.16%68.95%
2019-05-01TextMask_V170.58%67.33%68.92%
2019-04-22MFTD: Mask Filters for Text Detection63.05%72.09%67.27%
2021-04-23HOCRA64.35%69.75%66.94%
2019-04-25Art detect by vivo57.15%80.72%66.92%
2019-04-29PAT-S.Y59.64%75.72%66.72%
2019-04-16Art_test_baseline_task162.27%71.38%66.51%
2019-04-30DMCA64.01%69.08%66.45%
2019-04-30TMIS53.49%86.19%66.01%
2021-04-28NN_euro651.76%85.50%64.48%
2019-04-22mask rcnn55.61%74.83%63.81%
2019-05-01Unicamp-SRBR-PN-157.59%68.02%62.37%
2019-04-26TP51.62%78.18%62.18%
2019-04-28Improved Progressive scale expansion Net52.24%75.88%61.88%
2019-04-23159.04%57.38%58.20%
2019-04-27TextCohesion_143.66%68.08%53.20%
2019-04-30EM-DATA45.11%61.34%51.99%
2021-04-29HOCRA_base33.63%83.00%47.87%
2019-04-26RAST: Robust Arbitrary Shape Text Detector35.44%71.08%47.30%
2021-03-31inception baseline27.68%54.89%36.80%
2019-04-30MSR0.46%0.55%0.50%

Ranking Graphic