De-rendering Stylized Texts

Shimoda, Wataru; Haraguchi, Daichi; Uchida, Seiichi; Yamaguchi, Kota

doi:10.1109/iccv48922.2021.00111

Cited by 12 publications

(7 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Wu et al [7] proposed an end-to-end style retention network (SRNet) with text conversion, background inpainting, and fusion for editing text in the wild. Shimoda et al [8] suggested a text vectorization technique by leveraging the advantage of differentiable text rendering to accurately reproduce the input raster text in a resolutionfree parametric format. On the other hand, the printed character image generation [42]- [48] has been comprehensively studied.…”

Section: Text Image Generationmentioning

confidence: 99%

See 1 more Smart Citation

APRNet: Attention-based Pixel-wise Rendering Network for Photo-Realistic Text Image Generation

Shi¹,

Ding²,

Chen³

et al. 2022

Preprint

View full text Add to dashboard Cite

Style-guided text image generation tries to synthesize text image by imitating reference image's appearance while keeping text content unaltered. The text image appearance includes many aspects. In this paper, we focus on transferring style image's background and foreground color patterns to the content image to generate photo-realistic text image. To achieve this goal, we propose 1) a content-style cross attention based pixel sampling approach to roughly mimicking the style text image's background; 2) a pixel-wise style modulation technique to transfer varying color patterns of the style image to the content image spatial-adaptively; 3) a cross attention based multi-scale style fusion approach to solving text foreground misalignment issue between style and content images; 4) an image patch shuffling strategy to create style, content and ground truth image tuples for training. Experimental results on Chinese handwriting text image synthesis with SCUT-HCCDoc and CASIA-OLHWDB datasets demonstrate that the proposed method can improve the quality of synthetic text images and make them more photo-realistic.

show abstract

Section: Text Image Generationmentioning

confidence: 99%

“…Recently, neural network based style-guided approaches have been applied to printed [7], [8] and handwriting text image generation [1]- [3]. Equipped by the generative adversarial network (GAN) [9], style transfer [10]- [15] and imageto-image translation [16]- [19], the quality of synthetic text images has been improved greatly.…”

Section: Introductionmentioning

confidence: 99%

APRNet: Attention-based Pixel-wise Rendering Network for Photo-Realistic Text Image Generation

Shi¹,

Ding²,

Chen³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Taken a content image and a style image as inputs, text image editing (TIE) [45], [46], [47] aims to replace the text instance in the style image while retaining the styles of both the background and text. SRNet [45] first generated the foreground text by style transfer, then obtained the background image by text erasure, and finally synthesized the edited text image by fusing foreground and background.…”

Section: Text Image Editingmentioning

confidence: 99%

“…SwapText [46] further improved SRNet by manipulating geometric points of characters to transform text locations. Shimoda et al [47] formulated raster text editing as a de-rendering problem. They proposed a vectorization model to parse text information, and edited texts by a rendering engine using the parsed parameters.…”

Section: Text Image Editingmentioning

confidence: 99%

Unsupervised Structure-Texture Separation Network for Oracle Character Recognition

Wang,

Deng,

Liu

2022

Preprint

View full text Add to dashboard Cite

Oracle bone script is the earliest-known Chinese writing system of the Shang dynasty and is precious to archeology and philology. However, real-world scanned oracle data are rare and few experts are available for annotation which make the automatic recognition of scanned oracle characters become a challenging task. Therefore, we aim to explore unsupervised domain adaptation to transfer knowledge from handprinted oracle data, which are easy to acquire, to scanned domain. We propose a structure-texture separation network (STSN), which is an end-to-end learning framework for joint disentanglement, transformation, adaptation and recognition. First, STSN disentangles features into structure (glyph) and texture (noise) components by generative models, and then aligns handprinted and scanned data in structure feature space such that the negative influence caused by serious noises can be avoided when adapting. Second, transformation is achieved via swapping the learned textures across domains and a classifier for final classification is trained to predict the labels of the transformed scanned characters. This not only guarantees the absolute separation, but also enhances the discriminative ability of the learned features. Extensive experiments on Oracle-241 dataset show that STSN outperforms other adaptation methods and successfully improves recognition performance on scanned data even when they are contaminated by long burial and careless excavation.

show abstract

“…Besides, it can be widely used for document restoration in the field of intelligent education. It is also a crucial prerequisite step for text editing [49,52,53,20,37] and has wide applications in areas such as augmented reality translation. Recent text removal methods [56,23,48,38,40] have achieved significant improvements with the development of GAN [10,27,28].…”

Section: Introductionmentioning

confidence: 99%

Don't Forget Me: Accurate Background Recovery for Text Removal via Modeling Local-Global Context

Liu¹,

Jin²,

Liu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Text removal has attracted increasingly attention due to its various applications on privacy protection, document restoration, and text editing. It has shown significant progress with deep neural network. However, most of the existing methods often generate inconsistent results for complex background. To address this issue, we propose a Contextual-guided Text Removal Network, termed as CTRNet. CTRNet explores both low-level structure and high-level discriminative context feature as prior knowledge to guide the process of text erasure and background restoration. We further propose a Local-global Content Modeling (LGCM) block with CNNs and Transformer-Encoder to capture local features and establish the long-term relationship among pixels globally. Finally, we incorporate LGCM with context guidance for feature modeling and decoding. Experiments on benchmark datasets, SCUT-EnsText and SCUT-Syn show that CTRNet significantly outperforms the existing state-of-the-art methods. Furthermore, a qualitative experiment on examination papers also demonstrates the generalizability of our method. The code of CTRNet is available at https://github.com/lcy0604/CTRNet.

show abstract

De-rendering Stylized Texts

Cited by 12 publications

References 32 publications

APRNet: Attention-based Pixel-wise Rendering Network for Photo-Realistic Text Image Generation

APRNet: Attention-based Pixel-wise Rendering Network for Photo-Realistic Text Image Generation

Unsupervised Structure-Texture Separation Network for Oracle Character Recognition

Don't Forget Me: Accurate Background Recovery for Text Removal via Modeling Local-Global Context

Contact Info

Product

Resources

About