Turbo4DGen：面向 4D 生成的超快速加速框架
Turbo4DGen: Ultra-Fast Acceleration for 4D Generation
AI 导读
研究者提出 Turbo4DGen，这是首个专门面向扩散式多视角 4D 内容生成的加速框架。它通过跨去噪步骤复用中间注意力的时空缓存机制、语义感知注意力剪枝和自适应 SCM 链旁路调度器，大幅削减 spatio-camera-motion 注意力冗余计算。在 ObjaverseDy 和 Consistent4D 数据集上，Turbo4DGen 实现平均 9.7× 加速且质量无下降。
正文 · 原文
Abstract:4D generation, or dynamic 3D content generation, integrates spatial, temporal, and view dimensions to model realistic dynamic scenes, playing a foundational role in advancing world models and physical AI. However, maintaining long-chain consistency across both frames and viewpoints through the unique spatio-camera-motion (SCM) attention mechanism introduces substantial computational and memory overhead, often leading to out-of-memory (OOM) failures and prohibitive generation times. To address these challenges, we propose Turbo4DGen, an ultra-fast acceleration framework for diffusion-based multi-view 4D content generation. Turbo4DGen introduces a spatiotemporal cache mechanism that persistently reuses intermediate attention across denoising steps, combined with dynamically semantic-aware attention pruning and an adaptive SCM chain bypass scheduler, to drastically reduce redundant SCM attention computation. Our experimental results show that Turbo4DGen achieves an average 9.7$\times$ speedup without quality degradation on the ObjaverseDy and Consistent4D datasets. To the best of our knowledge, Turbo4DGen is the first dedicated acceleration framework for 4D generation.
| Comments: | ICML 2026 |
| Subjects: | Graphics (cs.GR); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2603.29572 [cs.GR] |
| (or arXiv:2603.29572v2 [cs.GR] for this version) | |
| https://doi.org/10.48550/arXiv.2603.29572 arXiv-issued DOI via DataCite |
Submission history
From: Yuanbin Man [view email]
[v1]
Sat, 24 Jan 2026 23:22:01 UTC (2,897 KB)
[v2]
Wed, 28 Oct 2026 06:07:08 UTC (2,761 KB)
来源：arXiv cs.AI · arxiv.org