English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Modeling sounds emitted from physical object interactions is critical forimmersive perceptual experiences in real and virtual worlds. Traditionalmethods of impact sound synthesis use physics simulation to obtain a set ofphysics parameters that could represent and synthesize the sound. However, theyrequire fine details of both the object geometries and impact locations, whichare rarely available in the real world and can not be applied to synthesizeimpact sounds from common videos. On the other hand, existing video-driven deeplearning-based approaches could only capture the weak correspondence betweenvisual content and impact sounds since they lack of physics knowledge. In thiswork, we propose a physics-driven diffusion model that can synthesizehigh-fidelity impact sound for a silent video clip. In addition to the videocontent, we propose to use additional physics priors to guide the impact soundsynthesis procedure. The physics priors include both physics parameters thatare directly estimated from noisy real-world impact sound examples withoutsophisticated setup and learned residual parameters that interpret the soundenvironment via neural networks. We further implement a novel diffusion modelwith specific training and inference strategies to combine physics priors andvisual information for impact sound synthesis. Experimental results show thatour model outperforms several existing systems in generating realistic impactsounds. More importantly, the physics-based representations are fullyinterpretable and transparent, thus enabling us to perform sound editingflexibly.

Physics-Driven Diffusion Models for Impact Sound Synthesis from Videos