MB-RIRs: a Synthetic Room Impulse Response Dataset with Frequency-Dependent Absorption Coefficients
This work addresses the need for more ecologically valid synthetic RIR datasets for speech enhancement researchers, though it is incremental as it builds on existing methods.
The paper tackled the problem of improving synthetic room impulse response datasets for monaural speech enhancement by implementing features like multiband absorption coefficients, finding that this approach achieved a +0.51dB SDR and +8.9 MUSHRA score when evaluated on real RIRs.
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.