COSY: 用于解耦人像编辑的组合式3DGS合成
COSY: Compositional 3DGS Synthesis for Disentangled Human Head Editing
AI总结：
提出一种组合式3DGS生成器架构，通过独立合成头发、皮肤、眼镜和躯干等组件实现语义属性的精确解耦编辑，无需分割掩码或几何先验。
摘要：
近期用于人像的3D高斯泼溅(3DGS) GAN能够实时合成和渲染逼真的3D模型，并在身份和外观上提供丰富的多样性。然而，控制特定语义属性（如发色或眼镜）仍然具有挑战性，因为纠缠的潜在空间中的编辑常常导致身份或外观的意外变化。尽管有几种方法旨在通过估计仅修改特定特征的方向来解耦训练后的潜在空间，但这些方法无法保证完全解耦，并且通常需要预训练的分类器。在我们的方法中，我们提出了一种新的生成器架构，该架构完全独立地合成组件，例如头发、皮肤、眼镜和躯干。这使得可以更改一个区域的潜在向量，同时保持其余部分固定。此外，我们仅使用稀疏信息（如头发或皮肤颜色）实现这种分离，消除了先前工作中常见的分割掩码或几何先验的需求。为了确保编辑过程中形状和光照条件的匹配，我们允许独立生成器之间通过上下文令牌共享最少的信息。这些令牌甚至使我们能够控制形状和光照，而无需任何先验注释。与现有的基于GAN的生成和编辑工作相比，我们的方法显示出更好的解耦性、更精确的编辑控制和有竞争力的视觉质量。
英文摘要：
Recent 3D Gaussian Splatting (3DGS) GANs for human heads synthesize and render photorealistic 3D models in real-time and offer a vast variety in identity and appearance. However, controlling specific semantic attributes such as hair color or glasses remains challenging, as edits in the entangled latent space often induce unintended changes in identity or appearance. Although there are several methods that aim to disentangle the latent space post training by estimating directions that only modify certain features, these methods cannot guarantee complete disentanglement and often require pre-trained classifiers. In our approach, we propose a new generator architecture that synthesizes components, such as hair, skin, glasses, and torso, completely independently. This allows for changing the latent vector for one region while keeping the remaining parts fixed. Further, we achieve this separation using only sparse information such as the hair or skin color, eliminating the requirement of segmentation masks or geometric priors, often seen in prior work. To ensure matching shape and lighting conditions during editing, we allow minimal shared information via context tokens between the independent generators. These tokens even allow us to control the shape and light, without any prior annotation. Compared to existing works on GAN-based generation and editing, our method shows better disentanglement, more precise editing control, and competitive visual quality.
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。
