Progressive Scene Text Erasing with Self-Supervision
This work improves text erasing in images for applications like privacy protection, but it is incremental as it builds on existing methods with a novel pretext task.
The paper tackles the problem of scene text erasing by addressing the gap between synthetic and real-world data, using self-supervision and a progressive network to improve generalization, achieving state-of-the-art performance on public benchmarks.
Scene text erasing seeks to erase text contents from scene images and current state-of-the-art text erasing models are trained on large-scale synthetic data. Although data synthetic engines can provide vast amounts of annotated training samples, there are differences between synthetic and real-world data. In this paper, we employ self-supervision for feature representation on unlabeled real-world scene text images. A novel pretext task is designed to keep consistent among text stroke masks of image variants. We design the Progressive Erasing Network in order to remove residual texts. The scene text is erased progressively by leveraging the intermediate generated results which provide the foundation for subsequent higher quality results. Experiments show that our method significantly improves the generalization of the text erasing task and achieves state-of-the-art performance on public benchmarks.