Authors: Xiufeng Jiang, Changhao Wu, Rongrong Wang, Zhiwei Jia, Yue Tao, Shugong Xu(advisor)

Description: 1. We use a novel text detector called Progressive Scale Expansion Network (PSENet) as a basic method from CVPR2019,which can precisely detect text instances with arbitrary shapes.More specifically, PSENet generates the different scale of kernels for each text instance, and gradually expands the minimal scale kernel to the text instance with the complete shape.Due to the fact that there are large geometrical margins among the minimal scale kernels, this method is effective to split the close text instances, making it easier to use segmentation-based methods to detect arbitrary-shaped text instances.
2. Scene text detection has witnessed rapid progress especially with the recent development of convolutional neural networks. However, there still exists two challenges which prevent the algorithm into industry applications. On the one hand, most of the state-of-art algorithms require quadrangle bounding box which is in-accurate to locate the texts with arbitrary shape. On the other hand, two text instances which are close to each other may lead to a false detection which covers both instances. Traditionally, the segmentation-based approach can relieve the first problem but usually fail to solve the second challenge.
3. We make some changes in the Psenet.First,we use Deformable Convolutional Networks(DCN v2) instead of traditional convolution in the backbone network(Resnet50).Second,we do some data enhancement on the dataset, especially in the frequency domain to do some changes.