Trained DeepONet models for predicting the sound field in 2D and 3D domains

We address the challenge of acoustic simulations in three-dimensional (3D) virtual rooms with parametric source positions, which have applications in virtual/augmented reality, game audio, and spatial computing. The wave equation can fully describe wave phenomena such as diffraction and interference. However, conventional numerical discretization methods are computationally expensive when simulating hundreds of source and receiver positions, making simulations with parametric source positions impractical. To overcome this limitation, we propose using deep operator networks to approximate linear wave-equation operators. This enables the rapid prediction of<br>sound propagation in realistic 3D acoustic scenes with parametric source positions, achieving millisecond-scale computations.The data in this repository consist of the trained DeepONet models for reproducing all results in the paper "Sound propagation in realistic interactive 3D scenes with parameterized sources using deep neural operators", by Borrel-Jensen et al. The trained models provided are for the cubic, L-shaped, furnished and dome geometries in 3D and for the transfer learning results in 2D.

本研究聚焦参数化声源位置下的三维（3D）虚拟房间声学仿真难题，此类场景在虚拟现实/增强现实、游戏音频及空间计算领域均具备重要应用价值。波动方程可完整刻画衍射、干涉等各类波动现象，但传统数值离散化方法在仿真数百个声源与接收点位置时计算成本高昂，导致参数化声源位置下的声学仿真难以实际落地。为突破这一局限，我们提出采用深度算子网络（DeepONet）对线性波动方程算子进行近似建模，由此可快速预测参数化声源位置下的真实三维声学场景中的声波传播过程，实现毫秒级的高效计算。本数据集仓库包含已训练完成的DeepONet模型，可复现Borrel-Jensen等人发表的学术论文《基于深度神经算子的参数化声源交互式真实三维场景声波传播》中的全部实验结果。所提供的训练模型覆盖三维场景中的立方体形、L形、带家具及穹顶四种几何结构，以及二维场景下的迁移学习实验结果。