Adaptive Shrink-Mask for Text Detection

TL;DR

The paper introduces an adaptive shrink-mask based text detection network that enhances accuracy and robustness while maintaining real-time speed, by weakening the dependence on shrink-masks and utilizing surrounding pixel information during training.

Contribution

It proposes the Adaptive Shrink-Mask (ASM) and Super-pixel Window (SPW) to improve text detection robustness and accuracy, with a lightweight architecture for efficiency.

Findings

Outperforms state-of-the-art methods in accuracy and speed.

Improves detection robustness by weakening coupling to shrink-masks.

Uses surrounding pixel context during training, not testing.

Abstract

Existing real-time text detectors reconstruct text contours by shrink-masks directly, which simplifies the framework and can make the model run fast. However, the strong dependence on predicted shrink-masks leads to unstable detection results. Moreover, the discrimination of shrink-masks is a pixelwise prediction task. Supervising the network by shrink-masks only will lose much semantic context, which leads to the false detection of shrink-masks. To address these problems, we construct an efficient text detection network, Adaptive Shrink-Mask for Text Detection (ASMTD), which improves the accuracy during training and reduces the complexity of the inference process. At first, the Adaptive Shrink-Mask (ASM) is proposed to represent texts by shrink-masks and independent adaptive offsets. It weakens the coupling of texts to shrink-masks, which improves the robustness of detection results. Then, the Super-pixel Window (SPW) is designed to supervise the network. It utilizes the surroundings of each pixel to improve the reliability of predicted shrink-masks and does not appear during testing. In the end, a lightweight feature merging branch is constructed to reduce the computational cost. As demonstrated in the experiments, our method is superior to existing state-of-the-art (SOTA) methods in both detection accuracy and speed on multiple benchmarks.