AES Fall Convention 2023

Deep learning/AI techniques have shown to be highly successful in audio signal processing, for example in denoising and audio source separation. In coding of audio signals there has been remarkably impressive performance for AI methods in coding speech at a low bit rate. Coding of general audio, such as music, has so far been a considerably harder challenge. However, recent advances in the field have resulted in AI-based systems that are competitive with traditional perceptual audio codecs.
This workshop, organized by the AES Technical Committee on Coding of Audio Signals, contains panelists from both industry and academia and will give an introduction to the topic, brief presentations of current techniques, followed by an interactive discussion between the audience and the panel about future directions and applications in this emerging field.