让语音唤醒支持用户自定义关键词和说话人验证，零样本部署
Personalized Keyword Spotting for User-Defined Keywords Leveraging Text-Independent Speaker Verification
- 用音素监督音频编码+预训练说话人编码，实现双零样本识别
- 在多个数据集上误拒率降低60%，参数仅155万适合边缘设备
- 无需重新训练即可切换普通唤醒或严格说话人认证模式
用户自定义关键词检测（UD-KWS）实现了基于文本的零样本唤醒词检测，但现有系统学习的是与说话人无关的表征，无法拒绝正确关键词但非本人说出的情况。本文提出ZP-KWS，一个轻量级框架，结合音素监督音频编码器与GE2E预训练的紧凑说话人编码器（约0.9M参数）。推理时采用乘法式晚期融合，使各分支拥有独立否决权，支持从常规检测到严格说话人门控激活的多种模式，无需重新训练。在LibriPhrase、Google Speech Commands和Qualcomm数据集上，ZP-KWS在1%误报率下目标仅误拒率相对基线最高降低60%，同时保持优异的关键词检测性能，整体参数量控制在155万以内，适用于边缘部署。
原文摘要 · Abstract (English)
User-defined keyword spotting (UD-KWS) enables zero-shot wake-word detection from text, but existing systems learn speaker-invariant representations that cannot reject impostors uttering the correct keyword. We address this dual zero-shot setting -- unseen keywords and unseen speakers -- with ZP-KWS, a lightweight framework combining a phoneme-supervised audio encoder with a GE2E-pretrained compact speaker encoder (about 0.9M parameters). Multiplicative late fusion at inference grants each branch independent veto power, supporting modes from conventional detection to strict speaker-gated activation without retraining. On LibriPhrase, Google Speech Commands, and Qualcomm datasets, ZP-KWS reduces target-only FRR at 1% FAR by up to 60% relative to the strongest baseline while maintaining competitive keyword detection, all within a 1.55M parameter budget for edge deployment.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。