Leveraging Spatial Cues from Cochlear Implant Microphones to Efficiently Enhance Speech Separation in Real-World Listening Scenes

2025-01-24

AI Summary

Cochlear implant (CI) users exhibit significantly degraded speech separation performance in realistic reverberant environments. Method: This paper proposes a robust speech separation method that synergistically integrates implicit and explicit spatial cues. We first systematically quantify the impact of real-world acoustic conditions on CI-based speech separation. To enhance spatial awareness—particularly under implicit-cue limitations (e.g., unilateral CI)—we design an explicit spatial feature auxiliary mechanism incorporating sound source localization, interaural time differences (ITDs), and interaural level differences (ILDs). Our deep learning architecture jointly optimizes end-to-end implicit representations with physically interpretable explicit features. Results: The method substantially improves separation quality in challenging scenarios involving adjacent speakers and spatially overlapping sources: it achieves a 2.3 dB improvement in SI-SNRi for single-channel CI input and notably enhances discriminability of phonetically similar speech.

Technology Category

• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Natural Language Processing: Speech
• Machine Learning: Multimodal Learning

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Speech separation approaches for single-channel, dry speech mixtures have significantly improved. However, real-world spatial and reverberant acoustic environments remain challenging, limiting the effectiveness of these approaches for assistive hearing devices like cochlear implants (CIs). To address this, we quantify the impact of real-world acoustic scenes on speech separation and explore how spatial cues can enhance separation quality efficiently. We analyze performance based on implicit spatial cues (inherent in the acoustic input and learned by the model) and explicit spatial cues (manually calculated spatial features added as auxiliary inputs). Our findings show that spatial cues (both implicit and explicit) improve separation for mixtures with spatially separated and nearby talkers. Furthermore, spatial cues enhance separation when spectral cues are ambiguous, such as when voices are similar. Explicit spatial cues are particularly beneficial when implicit spatial cues are weak. For instance, single CI microphone recordings provide weaker implicit spatial cues than bilateral CIs, but even single CIs benefit from explicit cues. These results emphasize the importance of training models on real-world data to improve generalizability in everyday listening scenarios. Additionally, our statistical analyses offer insights into how data properties influence model performance, supporting the development of efficient speech separation approaches for CIs and other assistive devices in real-world settings.

Problem

• Auditory Separation
• Cochlear Implants
• Complex Acoustic Environments

Innovation

• Spatial Information
• Speech Separation
• Real-world Data