Spatial LibriSpeech: An Augmented Dataset for Spatial Audio Learning
This provides a valuable resource for researchers in audio and machine learning, though it is incremental as it builds on existing LibriSpeech data with simulated augmentations.
The authors tackled the lack of large-scale spatial audio datasets for machine learning by creating Spatial LibriSpeech, an augmented dataset with over 650 hours of multi-channel audio and labels, which enabled training models that achieved a median absolute error of 6.60° on 3D source localization and similar gains on other tasks.
We present Spatial LibriSpeech, a spatial audio dataset with over 650 hours of 19-channel audio, first-order ambisonics, and optional distractor noise. Spatial LibriSpeech is designed for machine learning model training, and it includes labels for source position, speaking direction, room acoustics and geometry. Spatial LibriSpeech is generated by augmenting LibriSpeech samples with 200k+ simulated acoustic conditions across 8k+ synthetic rooms. To demonstrate the utility of our dataset, we train models on four spatial audio tasks, resulting in a median absolute error of 6.60° on 3D source localization, 0.43m on distance, 90.66ms on T30, and 2.74dB on DRR estimation. We show that the same models generalize well to widely-used evaluation datasets, e.g., obtaining a median absolute error of 12.43° on 3D source localization on TUT Sound Events 2018, and 157.32ms on T30 estimation on ACE Challenge.