LAG-Fusion: An Asynchronous Multimodal Diffusion Policy Fusion Framework
What happened
Zihao He and other authors propose LAG-Fusion, a delay-aware asynchronous multimodal diffusion policy composition framework: each modality's policy runs at its own inference rate and provides denoising guidance whenever it has output. To maintain consistency in asynchronous composition, the authors derive reference-frame rebasing rules for diffusion variables under relative action representations, so that guidance arriving late is aligned before fusion. The authors combine a low-frequency visual policy with a high-frequency force policy in contact-intensive manipulation. The authors state that experiments on heterogeneous modality delays show that LAG-Fusion outperforms synchronous fusion and a specially designed force-aware baseline in responsiveness and task performance.
AI 根据报道生成 · 2 days ago更新
Reporting timeline
Follow the reporting to explore different perspectives.
- arXiv cs.ROMATE: Asynchronous Multi-Rate Diffusion Policy Composition Framework
The authors propose LAG-Fusion, a latency-aware guidance fusion framework for asynchronous multimodal diffusion policy composition that allows modality-specific policies to run at their own inference rates and provide denoising guidance whenever output is available. To keep asynchronous composition consistent, the authors derive a reference-frame rebasing rule for diffusion variables under relative action representations, so that delayed guidance is aligned before fusion. In contact-rich manipulation, the authors compose a low-frequency vision policy with a high-frequency force policy; experiments with heterogeneous modality latency show better responsiveness and task performance than synchronous fusion and a purpose-designed force-aware baseline.
Activity for this event
- Current
- 3
- Peak
- 10Nov 9 14:00
- Past 24 hours
- -50%
The trend compares participants observed throughout the same period, so its range may be smaller than the current interest count. Move or tap to inspect each hour; use the arrow keys on a keyboard.