Adaptive Noise Resilient Keyword Spotting Using One-Shot Learning

2025-05-14

AI Summary

To address the significant degradation in keyword spotting (KWS) performance on resource-constrained embedded devices under dynamic acoustic noise, this paper proposes the first single-shot learning framework for noise-adaptive KWS. Methodologically, it leverages a lightweight pre-trained model and integrates noise-aware gradient regularization with a one-step fine-tuning strategy, enabling zero-shot incremental adaptation using only a single noisy utterance and one forward-backward pass. The approach achieves ultra-low latency (<10 ms), minimal memory overhead (<50 KB additional parameters), and strong robustness. Extensive experiments on real-world noise conditions (SNR ranging from 24 dB to −3 dB) demonstrate accuracy improvements of 4.9%–46.0% over baselines, with particularly pronounced gains under low-SNR regimes (≤18 dB). These results validate its suitability for real-time, on-device deployment in noisy edge environments.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Natural Language Processing: Learning & Optimization for NLP
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Keyword spotting (KWS) is a key component of smart devices, enabling efficient and intuitive audio interaction. However, standard KWS systems deployed on embedded devices often suffer performance degradation under real-world operating conditions. Resilient KWS systems address this issue by enabling dynamic adaptation, with applications such as adding or replacing keywords, adjusting to specific users, and improving noise robustness. However, deploying resilient, standalone KWS systems with low latency on resource-constrained devices remains challenging due to limited memory and computational resources. This study proposes a low computational approach for continuous noise adaptation of pretrained neural networks used for KWS classification, requiring only 1-shot learning and one epoch. The proposed method was assessed using two pretrained models and three real-world noise sources at signal-to-noise ratios (SNRs) ranging from 24 to -3 dB. The adapted models consistently outperformed the pretrained models across all scenarios, especially at SNR $leq$ 18 dB, achieving accuracy improvements of 4.9% to 46.0%. These results highlight the efficacy of the proposed methodology while being lightweight enough for deployment on resource-constrained devices.

Problem

• Improving keyword spotting robustness in noisy environments
• Enabling adaptive KWS with low computational resources
• Achieving high accuracy with one-shot learning adaptation

Innovation

• One-shot learning for noise adaptation
• Low computational continuous adaptation
• Lightweight deployment on resource-constrained devices