DPLM: A Deep Perceptual Spatial-Audio Localization Metric
This work addresses the need for efficient and cost-effective perceptual evaluation in audio-synthesis technologies like augmented and virtual reality, offering a novel metric to replace subjective tests.
The paper tackles the problem of capturing perceptual characteristics for localizing sounds by proposing a deep perceptual spatial-audio localization metric (DPLM) that assesses spatial localization differences between binaural recordings, and it outperforms baseline metrics in correlation with subjective ratings across diverse datasets without human-labeled training data.
Subjective evaluations are critical for assessing the perceptual realism of sounds in audio-synthesis driven technologies like augmented and virtual reality. However, they are challenging to set up, fatiguing for users, and expensive. In this work, we tackle the problem of capturing the perceptual characteristics of localizing sounds. Specifically, we propose a framework for building a general purpose quality metric to assess spatial localization differences between two binaural recordings. We model localization similarity by utilizing activation-level distances from deep networks trained for direction of arrival (DOA) estimation. Our proposed metric (DPLM) outperforms baseline metrics on correlation with subjective ratings on a diverse set of datasets, even without the benefit of any human-labeled training data.