解决扩散模型中罕见和小字文本渲染难题，提升生成准确性与视觉质量。
HDGlyph: A Hierarchical Disentangled Glyph-Based Framework for Long-Tail Text Rendering in Diffusion Models
- 分层解耦文本与图像生成，用字形网络分离字符特征。
- 英文和中文文本识别准确率分别提升5.08%和11.7%。
- 适合需要高精度长尾文本生成的设计、广告等场景。
视觉文本渲染旨在将指定文本内容准确融入生成图像，广泛应用于商业设计等领域。尽管近期取得进展，现有方法在处理罕见或小尺寸文本时仍表现不佳。本文提出一种分层解耦的字形基础框架HDGlyph，通过层级解耦文本生成与非文本视觉合成，实现对常见及长尾文本的联合优化。训练阶段，利用多语言字形网络（Multi-Linguistic GlyphNet）和字形感知感知损失（Glyph-Aware Perceptual Loss），在像素级上解耦表示，确保对未见字符的鲁棒渲染。推理阶段，采用噪声解耦无分类器引导与潜空间解耦双阶段渲染（LD-TSR）策略，分别优化背景与小尺寸文本。大量实验表明，本模型在英文和中文文本渲染上分别取得5.08%和11.7%的准确率提升，同时保持高质量图像输出，在长尾场景下也表现出色。
原文摘要 · Abstract (English)
Visual text rendering, which aims to accurately integrate specified textual content within generated images, is critical for various applications such as commercial design. Despite recent advances, current methods struggle with long-tail text cases, particularly when handling unseen or small-sized text. In this work, we propose a novel Hierarchical Disentangled Glyph-Based framework (HDGlyph) that hierarchically decouples text generation from non-text visual synthesis, enabling joint optimization of both common and long-tail text rendering. At the training stage, HDGlyph disentangles pixel-level representations via the Multi-Linguistic GlyphNet and the Glyph-Aware Perceptual Loss, ensuring robust rendering even for unseen characters. At inference time, HDGlyph applies Noise-Disentangled Classifier-Free Guidance and Latent-Disentangled Two-Stage Rendering (LD-TSR) scheme, which refines both background and small-sized text. Extensive evaluations show our model consistently outperforms others, with 5.08% and 11.7% accuracy gains in English and Chinese text rendering while maintaining high image quality. It also excels in long-tail scenarios with strong accuracy and visual performance.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。