Physics-Informed Transfer Learning for Data-Driven Sound Source Reconstruction in Near-Field Acoustic Holography

TL;DR

This paper introduces a transfer learning framework that combines data-driven neural networks with physics-informed fine-tuning to improve sound source reconstruction in near-field acoustic holography, enabling better generalization across different sound sources.

Contribution

It presents a novel two-stage transfer learning approach that integrates physics-informed fine-tuning with neural networks for improved sound source reconstruction in NAH.

Findings

Transfer learning improves reconstruction accuracy across different sound sources.

Physics-informed fine-tuning enhances model adaptation with minimal data.

The method achieves comparable or superior results to traditional techniques like C-ESM.

Abstract

We propose a transfer learning framework for sound source reconstruction in Near-field Acoustic Holography (NAH), which adapts a well-trained data-driven model from one type of sound source to another using a physics-informed procedure. The framework comprises two stages: (1) supervised pre-training of a complex-valued convolutional neural network (CV-CNN) on a large dataset, and (2) purely physics-informed fine-tuning on a single data sample based on the Kirchhoff-Helmholtz integral. This method follows the principles of transfer learning by enabling generalization across different datasets through physics-informed adaptation. The effectiveness of the approach is validated by transferring a pre-trained model from a rectangular plate dataset to a violin top plate dataset, where it shows improved reconstruction accuracy compared to the pre-trained model and delivers performance comparable to that of Compressive-Equivalent Source Method (C-ESM). Furthermore, for successful modes, the fine-tuned model outperforms both the pre-trained model and C-ESM in accuracy.