Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
The paper introduces Convex Language Detection (CLD), a convex optimization framework for robust language identification in low-resource speech recognition settings. CLD leverages multi-GPU ADMM in JAX for efficient training with global optimality guarantees and certified margin stability against feature perturbations. Experiments demonstrate that CLD achieves 97-98% accuracy in challenging low-resource scenarios, showcasing sample efficiency and robustness to dialectal variations.
Achieve 97-98% accuracy in low-resource language detection by reformulating the problem as a convex optimization, offering a surprisingly robust and efficient alternative to standard fine-tuning.
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/