Differentiable Acoustic Radiance Transfer

2025/09/19 by Lee, Sungho, Scerbo, Matteo, Han, Seungu +3
#Audio and Speech Processing (eess.AS) #FOS: Computer and information sciences #FOS: Electrical engineering #Signal Processing (eess.SP) #Sound (cs.SD) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2509.15946

Abstract

Geometric acoustics is an efficient approach to room acoustics modeling, governed by the canonical time-dependent rendering equation. Acoustic radiance transfer (ART) solves the equation through discretization, modeling the time- and direction-dependent energy exchange between surface patches given with flexible material properties. We introduce DART, a differentiable and efficient implementation of ART that enables gradient-based optimization of material properties. We evaluate DART on a simpler variant of the acoustic field learning task, which aims to predict the energy responses of novel source-receiver settings. Experimental results show that DART exhibits favorable properties, e.g., better generalization under a sparse measurement scenario, compared to existing signal processing and neural network baselines, while remaining a simple, fully interpretable system.

Citations

• Resounding Acoustic Fields with Reciprocity
• Differentiable physics for sound field reconstruction
• Time-domain sound field estimation using kernel ridge regression
• Differentiable Room Acoustic Rendering with Multi-View Vision Priors
• Hearing Anywhere in Any Environment
• Acoustic Volume Rendering for Neural Impulse Response Fields
• FLAMO: An Open-Source Library for Frequency-Domain Differentiable Audio Processing
• SOAF: Scene Occlusion-aware Neural Acoustic Field
• AV-GS: Learning Material and Geometry Aware Priors for Novel View Acoustic Synthesis
• Hearing Anything Anywhere
• NeRAF: 3D Scene Infused Neural Radiance and Acoustic Fields
• Real Acoustic Fields: An Audio-Visual Room Acoustics Dataset and Benchmark
• Room Acoustic Rendering Networks with Control of Scattering and Early Reflections
• Reconstruction of Sound Field through Diffusion Models
• A Review of Differentiable Digital Signal Processing for Music & Speech Synthesis
• AV-NeRF: Learning Neural Fields for Real-World Audio-Visual Scene Synthesis
• Novel-View Acoustic Synthesis
• Few-Shot Audio-Visual Learning of Environment Acoustics
• Learning Neural Acoustic Fields
• Differentiable Neural Radiosity
• Instant neural graphics primitives with a multiresolution hash encoding
• Differentiable Artificial Reverberation
• DDSP: Differentiable Digital Signal Processing
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• The Replica Dataset: A Digital Replica of Indoor Spaces
• Image method for efficiently simulating small-room acoustics
• Splitting the unit delay [FIR/all pass filters design]

Related

• Differentiable geometric acoustic path tracing using time-resolved path replay backpropagation
• Room Acoustic Rendering Networks with Control of Scattering and Early Reflections
• Modeling nonuniform energy decay through the modal decomposition of acoustic radiance transfer (MoD-ART)
• NeuralSound: Learning-based Modal Sound Synthesis With Acoustic Transfer
• NAT: Neural Acoustic Transfer for Interactive Scenes in Real Time