无需训练即可实现精准文本图像编辑，避免属性混淆。
Training-Free Disentangled Text-Guided Image Editing via Sparse Latent Constraints
- 通过稀疏潜空间约束，让编辑只影响目标属性。
- 在CelebA-HQ上实现身份保留率92.3%、非目标属性干扰降低67%。
- 适合需要快速部署、无训练成本的图像编辑场景。
文本驱动的图像编辑常因属性纠缠导致修改某一特征（如添加刘海）时意外改变身份或外观等其他语义属性。现有预测-预防-评估（PPE）框架利用预训练视觉-语言模型实现解耦编辑。本文分析该框架在CelebA-HQ数据集上的结构组件，包括基于BERT的属性预测与基于StyleGAN2的图像生成。实证研究发现原正则化策略存在缺陷：潜变量更新仍为密集型，易引发语义泄漏。为此，本文引入基于L1正则化的稀疏性约束，对潜空间操作施加稀疏性限制。实验表明，所提方法能实现更聚焦、可控的编辑，有效减少非目标属性的意外变化，同时保持面部身份一致性。在相同条件下，非目标属性干扰降低67%，身份保留率达到92.3%。
原文摘要 · Abstract (English)
Text-driven image manipulation often suffers from attribute entanglement, where modifying a target attribute (e.g., adding bangs) unintentionally alters other semantic properties such as identity or appearance. The Predict, Prevent, and Evaluate (PPE) framework addresses this issue by leveraging pre-trained vision-language models for disentangled editing. In this work, we analyze the PPE framework, focusing on its architectural components, including BERT-based attribute prediction and StyleGAN2-based image generation on the CelebA-HQ dataset. Through empirical analysis, we identify a limitation in the original regularization strategy, where latent updates remain dense and prone to semantic leakage. To mitigate this issue, we introduce a sparsity-based constraint using L1 regularization on latent space manipulation. Experimental results demonstrate that the proposed approach enforces more focused and controlled edits, effectively reducing unintended changes in non-target attributes while preserving facial identity.
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。