DynFOA conditions a latent diffusion model on 3D-Gaussian-splat geometry, material properties, and propagation features to synthesize first-order ambisonics from 360° video, reporting gains over OmniAudio and a new 600-clip benchmark — though the objective reference for the main benchmark is a fitte
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SD 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
DynFOA: Generating First-Order Ambisonics with Conditional Diffusion for Dynamic and Acoustically Complex 360-Degree Videos
DynFOA conditions a latent diffusion model on 3D-Gaussian-splat geometry, material properties, and propagation features to synthesize first-order ambisonics from 360° video, reporting gains over OmniAudio and a new 600-clip benchmark — though the objective reference for the main benchmark is a fitte