Low-Complexity Audio Embedding Extractors

Research output: Chapter in Book/Report/Conference proceedingConference proceedingspeer-review

Abstract

Solving tasks such as speaker recognition, music classification, or semantic audio event tagging with deep learning models typically requires computationally demanding networks. General-purpose audio embeddings (GPAEs) are dense representations of audio signals that allow lightweight, shallow classifiers to tackle various audio tasks. The idea is that a single complex feature extractor would extract dense GPAEs, while shallow MLPs can produce task-specific predictions. If the extracted dense representations are general enough to allow the simple downstream classifiers to generalize to a variety of tasks in the audio domain, a single costly forward pass suffices to solve multiple tasks in parallel. In this work, we try to reduce the cost of GPAE extractors to make them suitable for resource-constrained devices. We use efficient MobileNets trained on AudioSet using Knowledge Distillation from a Transformer ensemble as efficient GPAE extractors. We explore how to obtain high-quality GPAEs from the model, study how model complexity relates to the quality of extracted GPAEs, and conclude that low-complexity models can generate competitive GPAEs, paving the way for analyzing audio streams on edge devices w.r.t. multiple audio classification and recognition tasks.
Original languageEnglish
Title of host publicationProceedings of the European Signal Processing Conference (EUSIPCO), 2023
Number of pages5
Publication statusPublished - May 2023

Fields of science

  • 202002 Audiovisual media
  • 102 Computer Sciences
  • 102001 Artificial intelligence
  • 102003 Image processing
  • 102015 Information systems

JKU Focus areas

  • Digital Transformation

Cite this