30+ Years of Source Separation Research: Achievements and Future Challenges

2025-01-21

AI Summary

Sound source separation (SSS) has evolved significantly over three decades, yet a comprehensive, cross-domain historical analysis remains lacking. Method: This paper conducts a systematic literature review of SSS advancements from the 1990s to the 2020s, covering speech, music, and general audio in both single- and multi-channel settings. It integrates bibliometric analysis, critical evaluation of benchmarking frameworks (e.g., BSS_EVAL, MUSDB, LibriMix), and longitudinal trend analysis. Contribution/Results: We propose the first unified, cross-paradigm developmental taxonomy—comprising four stages: Blind Source Separation → Deep Learning → Interpretable Modeling → Embodied Interaction. Crucially, we identify how evaluation culture—encompassing metrics, datasets, and challenges—has actively shaped technical trajectories. We define three core challenges for next-generation SSS: causal interpretability, low-resource generalization, and human–machine collaborative separation. This work establishes an authoritative domain benchmark for ICASSP’s 50th anniversary and informs the transition toward post-deep-learning research paradigms.

Technology Category

• Machine Learning: Semi-Supervised Learning
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Models for Web evolution
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Source separation (SS) of acoustic signals is a research field that emerged in the mid-1990s and has flourished ever since. On the occasion of ICASSP's 50th anniversary, we review the major contributions and advancements in the past three decades in the speech, audio, and music SS research field. We will cover both single- and multi-channel SS approaches. We will also look back on key efforts to foster a culture of scientific evaluation in the research field, including challenges, performance metrics, and datasets. We will conclude by discussing current trends and future research directions.

Problem

• Source Separation
• Speech and Music Separation
• Future Challenges in Acoustics

Innovation

• Source Separation
• Audio Processing
• Machine Learning