MB-RIRs: a Synthetic Room Impulse Response Dataset with Frequency-Dependent Absorption Coefficients

TL;DR

This paper introduces MB-RIRs, a synthetic room impulse response dataset with frequency-dependent absorption coefficients, improving speech enhancement performance and matching real RIRs more closely.

Contribution

The paper presents a novel synthetic RIR dataset incorporating frequency-dependent absorption, source and receiver directivity, enhancing ecological validity for speech enhancement tasks.

Findings

MB-RIRs achieve +0.51dB SDR improvement on real RIRs

MB-RIRs obtain +8.9 MUSHRA score over traditional methods

The dataset is publicly available for research use

Abstract

We investigate the effects of four strategies for improving the ecological validity of synthetic room impulse response (RIR) datasets for monoaural Speech Enhancement (SE). We implement three features on top of the traditional image source method-based (ISM) shoebox RIRs: multiband absorption coefficients, source directivity and receiver directivity. We additionally consider mesh-based RIRs from the SoundSpaces dataset. We then train a DeepFilternet3 model for each RIR dataset and evaluate the performance on a test set of real RIRs both objectively and subjectively. We find that RIRs which use frequency-dependent acoustic absorption coefficients (MB-RIRs) can obtain +0.51dB of SDR and a +8.9 MUSHRA score when evaluated on real RIRs. The MB-RIRs dataset is publicly available for free download.