Styl3R：面向任意场景与风格的即时3D风格化重建
Styl3R: Instant 3D Stylized Reconstruction for Arbitrary Scenes and Styles
AI总结：
本文提出Styl3R，一种基于分支前馈架构和恒等损失预训练的即时3D风格化重建方法，可利用无位姿稀疏视角与任意风格图像在一秒内生成高一致性、高质量内容。
摘要：
在保持多视角一致性并忠实再现风格图像的同时，即时对3D场景进行风格化，仍然是一项重大挑战。当前最先进的3D风格化方法通常涉及计算密集型的测试时优化，以将艺术特征迁移到预训练的3D表示中，且往往需要密集的带位姿输入图像。相比之下，借助前馈重建模型的最新进展，我们展示了一种新颖方法：使用无位姿的稀疏视角场景图像和任意风格图像，在不到一秒内实现直接3D风格化。为解决重建与风格化之间固有的解耦问题，我们引入一种分支架构，将结构建模与外观着色分离，有效防止风格迁移扭曲底层3D场景结构。此外，我们调整了恒等损失，以通过新视角合成任务促进风格化模型的预训练。该策略还使模型在为风格化进行微调时保留其原有重建能力。使用域内和域外数据集进行的综合评估表明，我们的方法能够生成高质量的风格化3D内容，实现风格与场景外观的更优融合，并在多视角一致性和效率方面优于现有方法。
英文摘要：
Stylizing 3D scenes instantly while maintaining multi-view consistency and faithfully resembling a style image remains a significant challenge. Current state-of-the-art 3D stylization methods typically involve computationally intensive test-time optimization to transfer artistic features into a pretrained 3D representation, often requiring dense posed input images. In contrast, leveraging recent advances in feed-forward reconstruction models, we demonstrate a novel approach to achieve direct 3D stylization in less than a second using unposed sparse-view scene images and an arbitrary style image. To address the inherent decoupling between reconstruction and stylization, we introduce a branched architecture that separates structure modeling and appearance shading, effectively preventing stylistic transfer from distorting the underlying 3D scene structure. Furthermore, we adapt an identity loss to facilitate pre-training our stylization model through the novel view synthesis task. This strategy also allows our model to retain its original reconstruction capabilities while being fine-tuned for stylization. Comprehensive evaluations, using both in-domain and out-of-domain datasets, demonstrate that our approach produces high-quality stylized 3D content that achieve a superior blend of style and scene appearance, while also outperforming existing methods in terms of multi-view consistency and efficiency.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
