GISCA: Gradient-Inductive Segmentation Network With Contextual Attention for Scene Text Detection

Scene text detection (STD) is an irreplaceable step in a scene text reading system. It remains a more challenging task than general object detection since text objects are of arbitrary orientations and varying sizes. Generally, segmentation methods that use U-Net or hourglass-like networks are the mainstream approaches in multi-oriented text detection tasks. However, experience has shown that text-like objects in the complex background have high response values on the output feature map of U-Net, which leads to the severe false positive detection rate and degrades the STD performance. To tackle this issue, an adaptive soft attention mechanism called contextual attention module (CAM) is devised to integrate into U-Net to highlight salient areas and meanwhile retains more detail information. Besides, the gradient vanishing and exploding problems make U-Net more difficult to train because of the nonlinear deconvolution layer used in the up-sampling process. To facilitate the training process, a gradient-inductive module (GIM) is carefully designed to provide a linear bypass to make the gradient back-propagation process more stable. Accordingly, an end-to-end trainable Gradient-Inductive Segmentation network with Contextual Attention is proposed (GISCA). The experimental results on three public benchmarks have demonstrated that the proposed GISCA achieves the state-of-the-art results in terms of f-measure: 92.1%, 87.3%, and 81.4% for ICDAR 2013, ICDAR 2015, and MSRA TD500, respectively.