MorphAny3D：释放结构化潜变量在三维变形中的力量
MorphAny3D: Unleashing the Power of Structured Latent in 3D Morphing
提出在 3D 生成器的注意力机制中混合源与目标 SLAT 特征来实现无训练 3D 变形，其 Morphing Cross-Attention 和 Temporal-Fused Self-Attention 可直接迁移到其他基于 SLAT 的生成模型（如 TRELLIS）以支持时序平滑的编辑与风格迁移。
原文依据 intelligently blending source and target SLAT features within the attention mechanisms of 3D generators naturally produces plausible morphing sequences
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论三维变形仍然具有挑战性，因为难以生成语义一致且时序平滑的变形，尤其是跨类别时。我们提出 MorphAny3D，一个无需训练的框架，利用结构化潜变量（SLAT）表示实现高质量三维变形。我们的关键洞见是：在三维生成器的注意力机制内智能地混合源与目标的 SLAT 特征，能自然产生合理的变形序列。为此，我们引入变形交叉注意力（MCA），融合源与目标信息以实现结构连贯性；以及时序融合自注意力（TFSA），通过整合前序帧的特征来增强时序一致性。一种方向校正策略进一步缓解变形步骤中的姿态歧义。大量实验表明，我们的方法能生成最优的变形序列，即使在具有挑战性的跨类别情况下也是如此。MorphAny3D 还支持解耦变形和三维风格迁移等高级应用，并可推广到其他基于 SLAT 的生成模型。项目页面：https://xiaokunsun.github.io/MorphAny3D.github.io/。
英文原文摘要 1182 字符
3D morphing remains challenging due to the difficulty of generating semantically consistent and temporally smooth deformations, especially across categories. We present MorphAny3D, a training-free framework that leverages Structured Latent (SLAT) representations for high-quality 3D morphing. Our key insight is that intelligently blending source and target SLAT features within the attention mechanisms of 3D generators naturally produces plausible morphing sequences. To this end, we introduce Morphing Cross-Attention (MCA), which fuses source and target information for structural coherence, and Temporal-Fused Self-Attention (TFSA), which enhances temporal consistency by incorporating features from preceding frames. An orientation correction strategy further mitigates the pose ambiguity within the morphing steps. Extensive experiments show that our method generates state-of-the-art morphing sequences, even for challenging cross-category cases. MorphAny3D further supports advanced applications such as decoupled morphing and 3D style transfer, and can be generalized to other SLAT-based generative models. Project page: https://xiaokunsun.github.io/MorphAny3D.github.io/.