通过流模型修正潜在空间错位，提升图像生成的几何一致性。
Equivariant Latent Alignment via Flow Matching under Group Symmetries
- 基于流模型构建残差潜在空间，纠正对称性下的潜在错位
- 在SO(n)旋转群下显著降低潜在空间失配度，提升新视角合成质量
- 适合关注几何感知生成与对称性建模的研究者
几何感知生成模型和新视角合成方法在视觉保真度与一致性方面展现出巨大潜力。与此同时，等变表示学习作为一种强大框架，可在潜在空间中直接作用于已知的群变换，捕捉数据中的几何结构，提升新视角合成的可解释性与泛化能力。然而，我们发现现有方法常存在潜在空间错位问题，即预期的群作用与潜在空间实际所需的变换不一致，导致学习到的潜在变量无法一致保持底层对称性所施加的等变关系。为解决此问题，我们提出残差潜在流（Residual Latent Flow），一种基于流的框架，用于校正潜在空间中的错位，从而提高对底层等变关系的符合度。大量实验表明，该方法在旋转群SO(n)下显著减少潜在空间错位，并提升新视角合成质量。
原文摘要 · Abstract (English)
Geometry-aware generative models and novel view synthesis approaches have shown strong potential in visual fidelity and consistency. In parallel, equivariant representation learning has emerged as a powerful framework for constructing latent spaces where analytically known group transformations could act directly, capturing geometric structure in data and enhancing both interpretability and generalization in novel view synthesis. However, we identify that existing approaches often suffer from latent misalignment, a discrepancy between the intended group action and the actually required transformations in the latent space. Consequently, the learned latents often fail to consistently preserve the equivariant relations imposed by the underlying group symmetry. To address this, we propose Residual Latent Flow, a flow-based framework that corrects the misaligned latents, thereby improving compliance with the underlying equivariance relation. Our comprehensive experiments show that our method significantly reduces latent misalignment and improves novel view synthesis quality, under rotation groups SO(n).
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。