UDSS-BWE: Uncertainty- and Decision-Science Inspired Swin BandWidth Extension

2026-09-28

AI Summary

This study addresses the over-smoothing of high-frequency transient distortions in audio bandwidth extension by introducing decision science and uncertainty theory into this domain for the first time. Methodologically, it proposes a Swin Transformer-based dual-stream generator coupled with a learnable lattice technique, alongside five discriminators—including CVaRD and CCD—to enable complex-domain adversarial training. Experimental results on bilingual datasets demonstrate that the proposed approach significantly enhances perceptual quality while reducing the parameter count by 3.89 times, thereby establishing a new baseline for audio bandwidth extension.

Abstract

Bandwidth extension (BWE) is fundamentally localized: the most perceptual distortions are not average-case distortions, but rare high-frequency (HF) transients that standard, risk-neutral objectives tend to smooth away. To close this gap, we seek solutions in the risk-sensitive and uncertainty-aware decision science rules and present UDSS-BWE, which introduces five decision-science and uncertainty-aware discriminators: CVaRD (does tail pooling to amplify HF artifacts), CCD (a primal-dual augmented Lagrangian to prevent HF overboost), MCUD (a learnable utility over spectral flatness/ centroid/ rolloff), EDD (captures epistemic uncertainty), and DROD (captures entropic KL-DRO aggregation). UDSS-BWE is also designed as a complex valued adversarial BWE framework that uses Swin-based generators, a lightweight dual-stream shifted-window backbone, to capture local and long-range structure efficiently, while learnable lattice coupling provides controlled cross-stream exchange. UDSS-BWE is optimized extensively and achieves better perceptual quality with 3.89x fewer parameters (72M vs.18.5M) over two English and French datasets under clean and noisy conditions. To the best of our knowledge, this work shows how multi disciplinary decision-science-inspired and uncertainty theories can be successfully used to design efficient discriminators for producing more nuanced audios, establishing a new baseline in the BWE task.

Problem

• Bandwidth Extension
• High-Frequency Transients
• Perceptual Distortions
• Risk-Neutral Objectives
• Audio Quality

Innovation

• Bandwidth Extension
• Decision Science
• Uncertainty-aware Discriminators
• Swin Transformer
• Complex-valued Adversarial Framework