method: 360DigiTech_AI_OCR2022-06-06
Authors: Meichao Luo, Zhenming Wang, Tao Zhang, Qiang Wang, Xiaogang Huang
Affiliation: 360DigiTech_AI
Description: We proposed a method based on RARE and ensemble it with a modified CRNN model. Center Loss with initial value was also applied. Besides model modification, we use a huge public dataset for pretrain and generated lots of data with the train set image style for finetune. Finally, we use a language model to correct errors.
method: Linklogis_BeeAI2021-07-20
Authors: WeiJie Luo,Guopeng Wang, Keshu Chen,Shuyuan Chen, Zhiyuan Huang, Yan Xu
Affiliation: Linklogis_BeeAI Team
Description: We mainly completed Task 2 with a word-level text recognition approach, involving a specialized attention mechanism. In addition, we apply Central-Loss to improve accuracy.
method: BreSee OCR2021-07-06
Authors: Tianpeng Li, Linzhi Zhuang, Mengyue Shao, Jie Wu, Jiling Wu
Affiliation: BreSee AI Lab, Zhejiang Sci-Tech University
Description: We modify the SRN network structure and loss function, synthesize a large number of data sets, train SRN and fine tune on the training set, and use a large number of data enhancement operations.
Date | Method | Recall | Precision | Hmean | |||
---|---|---|---|---|---|---|---|
2022-06-06 | 360DigiTech_AI_OCR | 97.68% | 97.79% | 97.74% | |||
2021-07-20 | Linklogis_BeeAI | 97.67% | 97.69% | 97.68% | |||
2021-07-06 | BreSee OCR | 96.78% | 96.92% | 96.85% | |||
2021-04-01 | Linklogis_BigData | 96.48% | 96.89% | 96.68% | |||
2020-10-16 | Samsung Life Insurance | 96.68% | 96.46% | 96.57% | |||
2019-04-22 | H&H Lab | 96.35% | 96.52% | 96.43% | |||
2021-03-10 | Sunshine_OCR | 95.59% | 95.69% | 95.64% | |||
2022-08-02 | fff | 95.28% | 95.54% | 95.41% | |||
2021-01-16 | Sun Bear | 95.21% | 95.37% | 95.29% | |||
2021-02-19 | Hanoi University of Science and Technology - MSOLab | 95.05% | 95.04% | 95.04% | |||
2020-10-16 | Samsung Life Insurance | 95.04% | 94.82% | 94.93% | |||
2021-02-01 | DR Team | 94.63% | 95.17% | 94.90% | |||
2019-04-22 | HeReceipt-Ensemble | 94.56% | 95.10% | 94.82% | |||
2020-05-15 | Hanoi University of Science and Technology - MSOLab | 94.77% | 94.88% | 94.82% | |||
2019-04-22 | Ping An Property & Casualty Insurance Company | 94.48% | 94.86% | 94.67% | |||
2019-04-22 | CLOVA OCR | 94.30% | 94.88% | 94.59% | |||
2019-04-22 | SCUT-DLVC-Lab-Lexicon | 94.18% | 94.88% | 94.53% | |||
2019-04-22 | DenseNet-Attention Recognition | 94.29% | 94.58% | 94.44% | |||
2020-12-25 | hoainv | 94.31% | 94.34% | 94.33% | |||
2019-04-22 | CITlab Argus Text Recognition | 93.55% | 93.61% | 93.58% | |||
2021-01-03 | Sun Asterisk - AI Research | 93.47% | 93.58% | 93.52% | |||
2023-12-12 | ViTLP | 93.07% | 92.52% | 92.79% | |||
2019-12-19 | test | 90.60% | 90.70% | 90.65% | |||
2019-12-26 | test1 | 89.91% | 89.52% | 89.72% | |||
2020-11-27 | Hust | 90.02% | 88.66% | 89.34% | |||
2019-04-18 | Unet followed by CRNN with CTC | 88.58% | 87.30% | 87.93% | |||
2019-04-23 | BOE_IOT_AIBD | 87.84% | 86.66% | 87.24% | |||
2021-04-26 | GemAI - Use Resnet BiLSTM Attn | 84.06% | 90.63% | 87.22% | |||
2019-04-22 | CRNN after UNet Segmentation | 85.77% | 86.48% | 86.12% | |||
2019-04-22 | BiLstmCtcIgnoreSpaces-Segment | 83.38% | 87.37% | 85.33% | |||
2019-04-22 | IFLYTEK-textRec_v4 | 80.63% | 81.72% | 81.17% | |||
2019-04-22 | SituTech_OCR | 77.81% | 80.10% | 78.94% | |||
2020-10-14 | ResNet_BiLSTM_Attention | 74.05% | 77.81% | 75.88% | |||
2019-04-19 | Attention based OCR | 70.87% | 72.27% | 71.57% | |||
2021-04-14 | Practicing project for Scientific Research Subject (HCMUS master program) | 63.45% | 67.44% | 65.39% | |||
2019-04-18 | CTPN_CRNN | 35.75% | 63.89% | 45.85% | |||
2019-04-22 | A Text Extraction Method Based on Modified CRNN | 26.33% | 72.53% | 38.63% | |||
2019-04-18 | BiLSTM+ctc | 28.75% | 49.69% | 36.42% | |||
2019-04-23 | CRNN_Pytorch_BiCTC | 21.43% | 45.65% | 29.16% | |||
2019-04-21 | ICA-IVA | 19.48% | 40.32% | 26.27% | |||
2019-04-22 | Task 2 - Scanned Receipt OCR (Submitted by Intuit Inc.) | 1.05% | 3.96% | 1.66% | |||
2019-04-22 | Receipt Info Extracting Task2 letter-recognizing | 0.24% | 0.97% | 0.38% | |||
2019-04-16 | VIL | 0.00% | 0.00% | 0.00% | |||
2019-04-22 | CRNN_Autohome | 0.00% | 0.00% | 0.00% |