Spatial Audio
Overview
Spatial audio studies how listeners and machines use sound’s direction, distance, motion, and reverberation, enabling immersive playback, source tracking, and sound-guided interaction. Core approaches include first- and higher-order Ambisonics, spherical-harmonic processing, microphone-array beamforming, and binaural rendering with head-related transfer functions (HRTFs). Neural fields and diffusion or flow models now support tasks such as room-response prediction, spatial upmixing, and scene-grounded audio generation, while audio-language models are being tested on localization and motion-reasoning tasks. A central challenge is making systems generalize from simulated rooms and fixed arrays to real environments while preserving perceptually important cues; benchmarks such as DCASE and LOCATA, alongside human listening tests, help expose gaps in both spatial accuracy and evaluation metrics.
Latest papers
118 papers21 papers in the last four weeks, down 19% on the four weeks before. 0.2% of all new papers.