MorphAny3D: Unleashing the Power of Structured Latent in 3D Morphing
This addresses the problem of high-quality 3D morphing for applications like animation and design, though it appears incremental as it builds on existing SLAT-based generative models.
The paper tackles the challenge of generating semantically consistent and temporally smooth 3D morphing sequences, especially across categories, by introducing MorphAny3D, a training-free framework that leverages Structured Latent (SLAT) representations, achieving state-of-the-art results in experiments.
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/.