Authors: Jia Yu, Xinyu Zhou, Cong Yao, Jianan Wu, Chi Zhang, Shuchang Zhou

Description: The detection part is accomplished by a FCN which directly extracts text regions from original images.