Low-Rank Adaptation of Deep Prior Neural Networks For Room Impulse Response Reconstruction

2025-07-13

AI Summary

Deep priors suffer from poor generalization and limited adaptability to novel acoustic source positions in sound field reconstruction. To address this, we propose integrating Low-Rank Adaptation (LoRA) into the MultiResUNet deep prior framework for efficient, transferable room impulse response reconstruction from sparse sound pressure measurements. Our method avoids full-network retraining, enabling rapid adaptation to new acoustic configurations via updates to only a small set of low-rank parameters. Experiments demonstrate that LoRA maintains high physical fidelity—even with extremely few microphones (e.g., ≤4)—outperforming both conventional retraining and full-parameter fine-tuning. It achieves a 3.2× speedup in inference, introduces less than 0.5% additional parameters, and significantly reduces computational overhead while markedly enhancing cross-configuration generalization capability.

Technology Category

• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

The Deep Prior framework has emerged as a powerful generative tool which can be used for reconstructing sound fields in an environment from few sparse pressure measurements. It employs a neural network that is trained solely on a limited set of available data and acts as an implicit prior which guides the solution of the underlying optimization problem. However, a significant limitation of the Deep Prior approach is its inability to generalize to new acoustic configurations, such as changes in the position of a sound source. As a consequence, the network must be retrained from scratch for every new setup, which is both computationally intensive and time-consuming. To address this, we investigate transfer learning in Deep Prior via Low-Rank Adaptation (LoRA), which enables efficient fine-tuning of a pre-trained neural network by introducing a low-rank decomposition of trainable parameters, thus allowing the network to adapt to new measurement sets with minimal computational overhead. We embed LoRA into a MultiResUNet-based Deep Prior model and compare its adaptation performance against full fine-tuning of all parameters as well as classical retraining, particularly in scenarios where only a limited number of microphones are used. The results indicate that fine-tuning, whether done completely or via LoRA, is especially advantageous when the source location is the sole changing parameter, preserving high physical fidelity, and highlighting the value of transfer learning for acoustics applications.

Problem

• Adapting Deep Prior networks to new acoustic configurations efficiently
• Reducing computational cost of retraining for source position changes
• Enhancing generalization with Low-Rank Adaptation for sparse measurements

Innovation

• Uses Low-Rank Adaptation (LoRA) for efficient fine-tuning
• Embeds LoRA into MultiResUNet-based Deep Prior model
• Enables adaptation to new acoustic configurations quickly