Simulation-Based Inference for Plate Reverb System Identification

2026-09-28

AI Summary

This study addresses the high computational cost associated with inverting physical parameters from reverberation plate impulse responses. To overcome this challenge, we propose a simulation-based inference (SBI) framework that leverages neural networks to estimate posterior parameter distributions. Furthermore, a joint sampling and fine-tuning strategy is introduced to enable efficient inference without repeatedly invoking the simulator. This method effectively solves Task A of the DAFx Challenge, significantly improving both the accuracy and efficiency of parameter estimation while circumventing expensive forward simulations. By eliminating the need for iterative forward modeling, the proposed approach establishes a scalable new paradigm for acoustic system identification.

Abstract

We address Task A of the 1st DAFx Parameter Estimation Challenge, which aims to retrieve the physical parameters of a plate model from an impulse response. To do so, we use the Simulation-Based Inference (SBI) framework, in which we train a neural network to estimate a density over plate parameters given an impulse response, using a dataset generated by the simulator. Inference for a new impulse response then requires only a forward pass through the network, without involving the simulator. For each test observation, we fine-tune a specific network: additional simulation rounds are performed by sampling parameters from the current estimated distribution, simulating the corresponding impulse responses, and fine-tuning to produce the specialized network.

Problem

• Plate Reverb
• System Identification
• Parameter Estimation
• Impulse Response

Innovation

• Simulation-Based Inference
• Plate Reverb
• Parameter Estimation
• Neural Density Estimation
• Adaptive Fine-tuning