arXiv：cs.AI· Peiwei Chang, Yongqiang Chen, Yuelan Cheng, Liting Gao, Yishun Lu, Li Zhang, Mark D. Plumbley, Philip J. B. Jackson, Wenwu Wang·· 9 小时前
PhysicsSliders：为音频拟音生成学习物理控制
PhysicsSliders: Learning Physical Controls for Audio Foley Generation
AI 导读
PhysicsSliders（PXS）面向交互物体音效生成，用15个材质、几何和接触参数在模拟接触音上训练，提供直接参数输入与数值描述两种变体，可控制硬度、尺寸、速度与阻尼并支持房间描述。联合控制阻尼与尺寸时指定厚度可提升频率准确度；在五个真实音频数据集的18类撞击上，PXS-Physical 平均分组 Spearman 相关为 0.976，类中位误差 0.094 dex。
正文
Abstract:Physical properties provide intuitive controls for sounds made by interacting objects. The challenge is to make a generative model respond predictably to these controls, both individually and in combination. We present \method\ (PXS), trained on simulated contact sounds with 15 material, geometry and contact parameters. Two separately trained variants accept direct parameter inputs or numerical captions. PXS controls hardness, size, speed and damping, supports room descriptions and produces spectrally distinct outputs at fixed targets. When controlling damping and size together, specifying thickness improves frequency accuracy and joint target delivery. Measurements from real recordings define effective-damping targets, which simulator inversion maps to generation inputs without real-audio fine-tuning. Across 18 impact classes from five real-audio datasets, PXS-Physical follows these targets with mean grouped Spearman correlation $\rho=0.976$ and class-median error 0.094 dex (a factor of 1.24).
| Subjects: | Sound (cs.SD); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.18597 [cs.SD] |
| (or arXiv:2610.18597v1 [cs.SD] for this version) | |
| https://doi.org/10.48550/arXiv.2610.18597 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Peiwei Chang [view email]
[v1]
Tue, 13 Oct 2026 15:53:45 UTC (156 KB)
来源：arXiv：cs.AI · arxiv.org