# Self-Supervised Text Erasing with Controllable Image Synthesis

arXiv: https://arxiv.org/abs/2204.12743 · HTML: https://mlsanity.com/paper/2204.12743/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 35 of its 55 references that are in the map.
- First posted on arXiv 27 Apr 2022.

## Builds on (most cited first, 30 of 35)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Image-to-Image Translation with Conditional Adversarial Networks](https://mlsanity.com/paper/1611.07004/) (2016)
- [Domain-Adversarial Training of Neural Networks](https://mlsanity.com/paper/1505.07818/) (2015)
- [Large Scale GAN Training for High Fidelity Natural Image Synthesis](https://mlsanity.com/paper/1809.11096/) (2018)
- [V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation](https://mlsanity.com/paper/1606.04797/) (2016)
- [RandAugment: Practical automated data augmentation with a reduced search space](https://mlsanity.com/paper/1909.13719/) (2019)
- [Conditional Generative Adversarial Nets](https://mlsanity.com/paper/1411.1784/) (2014)
- [Neural Architecture Search with Reinforcement Learning](https://mlsanity.com/paper/1611.01578/) (2016)
- [Learning Transferable Features with Deep Adaptation Networks](https://mlsanity.com/paper/1502.02791/) (2015)
- [CyCADA: Cycle-Consistent Adversarial Domain Adaptation](https://mlsanity.com/paper/1711.03213/) (2017)
- [Least Squares Generative Adversarial Networks](https://mlsanity.com/paper/1611.04076/) (2016)
- [Playing for Data: Ground Truth from Computer Games](https://mlsanity.com/paper/1608.02192/) (2016)
- [Maximum Classifier Discrepancy for Unsupervised Domain Adaptation](https://mlsanity.com/paper/1712.02560/) (2017)
- [Deep Transfer Learning with Joint Adaptation Networks](https://mlsanity.com/paper/1605.06636/) (2016)
- [Generalizing to Unseen Domains via Adversarial Data Augmentation](https://mlsanity.com/paper/1805.12018/) (2018)
- [Synthetic Data for Text Localisation in Natural Images](https://mlsanity.com/paper/1604.06646/) (2016)
- [Certifying Some Distributional Robustness with Principled Adversarial Training](https://mlsanity.com/paper/1710.10571/) (2017)
- [Bridging Theory and Algorithm for Domain Adaptation](https://mlsanity.com/paper/1904.05801/) (2019)
- [Character Region Awareness for Text Detection](https://mlsanity.com/paper/1904.01941/) (2019)
- [Domain Randomization and Pyramid Consistency: Simulation-to-Real Generalization without Accessing Target Domain Data](https://mlsanity.com/paper/1909.00889/) (2019)
- [Synthetic Data and Artificial Neural Networks for Natural Scene Text Recognition](https://mlsanity.com/paper/1406.2227/) (2014)
- [Image to Image Translation for Domain Adaptation](https://mlsanity.com/paper/1712.00479/) (2017)
- [PP-OCR: A Practical Ultra Lightweight OCR System](https://mlsanity.com/paper/2009.09941/) (2020)
- [ICDAR2019 Robust Reading Challenge on Arbitrary-Shaped Text (RRC-ArT)](https://mlsanity.com/paper/1909.07145/) (2019)
- [Synthetic Data Augmentation using GAN for Improved Liver Lesion Classification](https://mlsanity.com/paper/1801.02385/) (2018)
- [Adversarial AutoAugment](https://mlsanity.com/paper/1912.11188/) (2019)
- [Editing Text in the Wild](https://mlsanity.com/paper/1908.03047/) (2019)
- [PointAugment: an Auto-Augmentation Framework for Point Cloud Classification](https://mlsanity.com/paper/2002.10876/) (2020)
- [SwapText: Image Based Texts Transfer in Scenes](https://mlsanity.com/paper/2003.08152/) (2020)
- [Verisimilar Image Synthesis for Accurate Detection and Recognition of Texts in Scenes](https://mlsanity.com/paper/1807.03021/) (2018)

## Built on by (most cited first, 9 of 9)

- [PosterLayout: A New Benchmark and Approach for Content-aware Visual-Textual Presentation Layout](https://mlsanity.com/paper/2303.15937/) (2023)
- [Visual Text Processing: A Comprehensive Review and Unified Evaluation](https://mlsanity.com/paper/2504.21682/) (2025)
- [TextPainter: Multimodal Text Image Generation with Visual-harmony and Text-comprehension for Poster Design](https://mlsanity.com/paper/2308.04733/) (2023)
- [ViTEraser: Harnessing the Power of Vision Transformers for Scene Text Removal with SegMIM Pretraining](https://mlsanity.com/paper/2306.12106/) (2023)
- [UPOCR: Towards Unified Pixel-Level OCR Interface](https://mlsanity.com/paper/2312.02694/) (2023)
- [Visual Text Meets Low-level Vision: A Comprehensive Survey on Visual Text Processing](https://mlsanity.com/paper/2402.03082/) (2024)
- [Orientation-Independent Chinese Text Recognition in Scene Images](https://mlsanity.com/paper/2309.01081/) (2023)
- [ConText: Driving In-context Learning for Text Removal and Segmentation](https://mlsanity.com/paper/2506.03799/) (2025)
- [Leveraging Text Localization for Scene Text Removal via Text-aware Masked Image Modeling](https://mlsanity.com/paper/2409.13431/) (2024)

## Abstract (arXiv)

Recent efforts on scene text erasing have shown promising results. However, existing methods require rich yet costly label annotations to obtain robust models, which limits the use for practical applications. To this end, we study an unsupervised scenario by proposing a novel Self-supervised Text Erasing (STE) framework that jointly learns to synthesize training images with erasure ground-truth and accurately erase texts in the real world. We first design a style-aware image synthesis function to generate synthetic images with diverse styled texts based on two synthetic mechanisms. To bridge the text style gap between the synthetic and real-world data, a policy network is constructed to control the synthetic mechanisms by picking style parameters with the guidance of two specifically designed rewards. The synthetic training images with erasure ground-truth are then fed to train a coarse-to-fine erasing network. To produce better erasing outputs, a triplet erasure loss is designed to enforce the refinement stage to recover background textures. Moreover, we provide a new dataset (called PosterErase), which contains 60K high-resolution posters with texts and is more challenging for the text erasing task. The proposed method has been extensively evaluated with both PosterErase and the widely-used SCUT-Enstext dataset. Notably, on PosterErase, our unsupervised method achieves 5.07 in terms of FID, with a relative performance of 20.9% over existing supervised baselines.