method: 4Paradigm-Data-Intelligence2019-06-03

Authors: Feng Cheng, Lixin Gu, Qingjie Liu, Feng Han, Jingtao Han

Description: The detection model and recognition model are trained separately.
Detection model: Based on Mask-RCNN. multi-scale. Train-set: 2017 MLT task1 train-set.
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: DeepSolo++ (ResNet-50)2023-05-22

Authors: Maoyuan Ye

Description: DeepSolo++ (Res-50, routing, #3)

@inproceedings{ye2022deepsolo, title={DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting}, author={Ye, Maoyuan and Zhang, Jing and Zhao, Shanshan and Liu, Juhua and Liu, Tongliang and Du, Bo and Tao, Dacheng}, booktitle={CVPR}, year={2023} }

method: PMTD + CNN based method2019-09-28

Authors: Geonho Hwang

Affiliation: NCIA, Seoul National University

Description: Task1: PMTD
Task2: CNN based method

Ranking Table

Description Paper Source Code
DateMethodHmeanPrecisionRecallAverage Precision
2023-05-22DeepSolo++ (ResNet-50)74.11%84.61%65.93%63.34%
2019-09-28PMTD + CNN based method72.64%78.76%67.39%61.48%

Ranking Graphic

Ranking Graphic