Authors: Wei Wang, Jin Wei, Chengxiang Ran, Xinye Yang, Tianjiao Cao, Fangmin Zhao, Yu Zhou

Affiliation: Institute of Information Engineering, Chinese Academy of Sciences; Mashang Consumer Finance Co., Ltd

Description: End-to-end text spotting method based on DETR. Trained with Synthtext-Chinese, ReCTs, LSVT, ART, 100k synthtic seal images and train set. The synthtic images are generated by taking WTW dataset(https://github.com/wangwen-whu/WTW-Dataset/) as background images and Chinese Company-Names-Corpus(https://github.com/wainshine/Company-Names-Corpus) as corpus.