Research
GeoCache: Training-Free Acceleration of Multi-View Texture Diffusion via Geometric Delta Transport
arXiv:2608.13255v1 Announce Type: cross Abstract: Geometry-conditioned multi-view diffusion enables high-quality 3D texture generation, but its repeated per-view denoiser evaluations introduce substan
arXiv:2608.13255v1 Announce Type: cross Abstract: Geometry-conditioned multi-view diffusion enables high-quality 3D texture generation, but its repeated per-view denoiser evaluations introduce substantial computational cost. Existing training-free accelerators primarily exploit temporal redundancy by reusing computation across denoising steps. In multi-view texturing, however, skipping a step also removes the cross-view interaction that continually aligns different observations of the same surface, leading to rapidly degraded consistency and fidelity. Our analysis identifies a complementary source of redundancy: although intermediate features remain view-specific, geometrically corresponding surface points exhibit transferable evolution in their predicted clean signals. Based on this observation, we introduce gc{}, a training-free plugin that evaluates a rotating subset of anchor views and transports their geometry-aligned per-step xz updates to the remaining views. Periodic full-view computation controls accumulated error, while sampler-consistent reconstruction preserves the denoising trajectory. gc{} requires neither retraining nor architectural modification and uses the position maps already available in geometry-conditioned texturing pipelines. Across Hunyuan3D-2.1, SyncMVD, and MVPainter, gc{} achieves a stronger speed--fidelity trade-off than temporal caches and step reduction at operating points above 2imes. On Hunyuan3D-2.1, it delivers a 2.21imes denoiser-loop speedup with an MV-LPIPS of 0.0293 and an MV-PSNR of 33.60 dB, providing the best fidelity among all tested methods above 2imes. The same transferred configuration reaches the highest speedup and lowest FLOPs on SyncMVD, while gc{} achieves the lowest FLOPs and best fidelity among the accelerated methods on MVPainter. These results establish cross-view geometry as an effective acceleration axis for multi-view texture diffusion.
Related
- HASTE: Training-Free Video Diffusion Acceleration via Head-Wise Adaptive Sparse Attention
- EO-VGGT: Orbital Ray-Conditioned 3D Foundation Models for Satellite Multi-View Reconstruction
- LoSA: Near-Lossless Sparse Attention for Training-Free Video Diffusion Acceleration
Source: arXiv cs.AI | 2026-08-14