FlexAvatar：面向细节形变可动画高斯头部化身的灵活大重建模型
FlexAvatar: Flexible Large Reconstruction Model for Animatable Gaussian Head Avatars with Detailed Deformation
用结构化头部 query token 作为规范锚点，把数量不定、无相机位姿、无表情标签的输入聚合为规范 3D 表示，再以 UV 空间位置图条件化的轻量 UNet 解码器实时生成表情相关形变；这套「规范锚点 + 条件形变解码」范式可直接迁移到其他可动画 3D 化身或动态人体重建任务。
原文依据 a transformer-based reconstruction model with structured head query tokens as a canonical anchor
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文提出 FlexAvatar，一种灵活的大重建模型，可从单张或稀疏图像重建高保真 3D 头部化身并生成细节丰富的动态形变，且无需相机位姿或表情标签。该方法采用基于 Transformer 的重建模型，以结构化头部查询 token 作为规范锚点，将数量不定、无相机位姿、无表情标签的输入聚合为鲁棒的规范 3D 表示。为刻画细节丰富的动态形变，本文引入以 UV 空间位置图为条件的轻量 UNet 解码器，可实时生成与表情相关的细节形变。为更好地捕捉皱纹、露齿等罕见但关键的表情，本文在训练中采用数据分布调整策略，以平衡这些表情在训练集中的分布。此外，仅需约 10 秒的轻量微调即可进一步增强极端身份下的身份特定细节，且不影响形变质量。大量实验表明，与已有方法相比，FlexAvatar 在 3D 一致性与动态细节真实感方面均取得更优表现，为可动画 3D 化身创建提供了实用方案。
英文原文摘要 1223 字符
We present FlexAvatar, a flexible large reconstruction model for high-fidelity 3D head avatars with detailed dynamic deformation from single or sparse images, without requiring camera poses or expression labels. It leverages a transformer-based reconstruction model with structured head query tokens as a canonical anchor to aggregate flexible input-number-agnostic, camera-pose-free and expression-free inputs into a robust canonical 3D representation. For detailed dynamic deformation, we introduce a lightweight UNet decoder conditioned on UV-space position maps, which can produce detailed expression-dependent deformations in real time. To better capture rare but critical expressions like wrinkles and bared teeth, we also adopt a data distribution adjustment strategy during training to balance the distribution of these expressions in the training set. Moreover, a lightweight 10-second refinement can further enhance identity-specific details in extreme identities without affecting deformation quality. Extensive experiments demonstrate that our FlexAvatar achieves superior 3D consistency, detailed dynamic realism compared with previous methods, providing a practical solution for animatable 3D avatar creation.