- Task 1 - Text Localization
- Task 2 - Script identification
- Task 3 - Joint text detection and script identification
method: Multi_scale_v22019-09-27
Authors: Wuheng Xu, Changxu Cheng, Bohan Li
Description: Area block feature information using images on multiple scales.This model has 4 scales and 8 branches.We used three training sets(mlt17, mlt19, mlt19val).
method: 4Paradigm-Data-Intelligence2019-05-30
Authors: ACVG
Description: Recognition model: Based on Transformer with backbone ResNet50. A voting process is done to identify the language of recognized transcript. Train-set: 2017 MLT task2 train-set & 2019 MLT task2 train-set & 2019 MLT Synthetic dataset.
method: Convolution based method 32019-09-27
Authors: Geonho Hwang
Affiliation: NCIA, Seoul National University
Description: This method is based on E2E-MLT-an Unconstrained End-to-End Method for Multi-Language Scene Text.
We use ICDAR 2017,2019 training, validation and synthetic data.
We ensemble entire 15 results from different networks, different training settings or different image scales.
| Date | Method | Script classification accuracy | |||
|---|---|---|---|---|---|
| 2019-09-27 | Multi_scale_v2 | 90.14% | |||
| 2019-05-30 | 4Paradigm-Data-Intelligence | 90.12% | |||
| 2019-09-27 | Convolution based method 3 | 89.60% | |||
| 2019-09-26 | Convolution based method 2 | 89.43% | |||
| 2019-01-19 | HUST(GSPA) | 89.42% | |||
| 2018-08-01 | CLOVA-AI / PAPAGO | 89.01% | |||
| 2019-01-17 | HUST(GS) | 88.51% | |||
| 2019-09-10 | CNN based method | 88.36% | |||
| 2017-07-02 | CNN based method 7 | 88.09% | |||
| 2017-06-28 | SCUT-DLVClab | 87.69% | |||
| 2017-07-01 | CNN based method 4 | 87.33% | |||
| 2017-07-01 | CNN based method 5 | 86.97% | |||
| 2017-06-30 | CNN based method 2 | 86.60% | |||
| 2017-07-02 | BLCT | 86.34% | |||
| 2017-07-02 | BLCT | 86.24% | |||
| 2017-06-02 | ecn-based method | 82.20% | |||
| 2017-07-01 | TH-DL | 80.72% | |||
| 2017-07-01 | An approach towards Word-Level Multi-Script Identification using Deep Transfer Features and SVM | 74.81% | |||
| 2017-07-01 | TNet | 48.33% | |||
| 2017-07-01 | TH-CNN | 43.22% |