Convex Low-resource Accent-Robust Language Detection in Speech Recognition

2026-05-22

AI Summary

This work addresses the challenge of inaccurate recognition of underrepresented accents and dialects in low-resource settings, which often leads to failure in downstream tasks. To this end, the authors propose Convex Language Detection (CLD), a novel framework that introduces, for the first time, convex optimization with theoretical guarantees into accent-robust language identification. Implemented in JAX, CLD leverages a multi-GPU Alternating Direction Method of Multipliers (ADMM) to efficiently converge to a globally optimal solution within polynomial time, while offering certified margin stability and robustness against adversarial feature perturbations. Experimental results demonstrate that CLD achieves 97–98% language identification accuracy under low-resource conditions, significantly enhancing robustness to accent variation and sample efficiency.

Abstract

Globalization and multiculturalism continue to produce increasingly diverse speech varieties. Yet current spoken dialogue systems frequently fail on under-represented dialects and accents, often misidentifying the input language and causing cascading failures in downstream dialogue tasks. Addressing this dialectal variance under low-resource constraints remains an open challenge, as standard fine-tuning is computationally expensive and prone to overfitting on high-dimensional speech data. We propose Convex Language Detection (CLD), a novel framework that integrates theoretically grounded convex optimization techniques into the spoken dialogue systems pipeline. Our method is efficiently implemented via multi-GPU Alternating Direction Method of Multipliers (ADMM) in JAX, thus providing global optimality guarantees and fast training in polynomial time. Theoretically, we prove that our convex objective induces certified margin stability and provide guarantees against feature perturbations. Empirically, we demonstrate sample efficiency and robustness to input dialectical variation, achieving 97-98% accuracy in challenging low-resource regimes. Our open-source package is available at https://pypi.org/project/jaxcld/

Problem

• accent-robust
• low-resource
• language detection
• dialectal variance
• speech recognition

Innovation

• convex optimization
• low-resource language detection
• accent robustness
• ADMM
• speech recognition