Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Sound Scene Synthesis at the DCASE 2024 Challenge

Mathieu Lagrange, Junwon Lee, Modan Tailleur, Laurie M. Heller, Keunwoo Choi, Brian McFee, Keisuke Imoto, Yuki Okamoto

TL;DR

This paper introduces a standardized evaluation framework for comparing different sound scene synthesis systems, incorporating both objective and subjective metrics, and reveals significant insights into the current capabilities and limitations of sound scene synthesis systems.

Abstract

This paper presents Task 7 at the DCASE 2024 Challenge: sound scene synthesis. Recent advances in sound synthesis and generative models have enabled the creation of realistic and diverse audio content. We introduce a standardized evaluation framework for comparing different sound scene synthesis systems, incorporating both objective and subjective metrics. The challenge attracted four submissions, which are evaluated using the Fréchet Audio Distance (FAD) and human perceptual ratings. Our analysis reveals significant insights into the current capabilities and limitations of sound scene synthesis systems, while also highlighting areas for future improvement in this rapidly evolving field.

Sound Scene Synthesis at the DCASE 2024 Challenge

TL;DR

This paper introduces a standardized evaluation framework for comparing different sound scene synthesis systems, incorporating both objective and subjective metrics, and reveals significant insights into the current capabilities and limitations of sound scene synthesis systems.

Abstract

This paper presents Task 7 at the DCASE 2024 Challenge: sound scene synthesis. Recent advances in sound synthesis and generative models have enabled the creation of realistic and diverse audio content. We introduce a standardized evaluation framework for comparing different sound scene synthesis systems, incorporating both objective and subjective metrics. The challenge attracted four submissions, which are evaluated using the Fréchet Audio Distance (FAD) and human perceptual ratings. Our analysis reveals significant insights into the current capabilities and limitations of sound scene synthesis systems, while also highlighting areas for future improvement in this rapidly evolving field.
Paper Structure (16 sections, 2 equations, 1 figure, 1 table)

This paper contains 16 sections, 2 equations, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Problem and Task Definition
  3. Dataset and Baseline
  4. Dataset Creation
  5. Sound Categories
  6. Baseline System
  7. Evaluation Methodology
  8. Objective Evaluation
  9. Subjective Evaluation
  10. Results
  11. System Performance
  12. Analysis
  13. Participation Analysis
  14. Discontinuation of the Task
  15. Conclusion
  16. ...and 1 more sections

Figures (1)

  • Figure 1: Correlation between FAD scores on evaluation set and other indicators, computed on the 4 submitted systems and the baseline system.