Smart audio devices are gated by an always-on lightweight keyword spotting program to reduce power consumption. It is however challenging to design models that have both high accuracy and low latency for accurate and fast responsiveness. Many efforts have been made to develop end-to-end neural networks, in which depthwise separable convolutions, temporal convolutions, and LSTMs are adopted as building units. Nonetheless, these networks designed with human expertise may not achieve an optimal trade-off in an expansive search space. In this paper, we propose to leverage recent advances in differentiable neural architecture search to discover more efficient networks. Our found model attains 97.2% top-1 accuracy on Google Speech Command Dataset v1.

中文翻译：

---

AutoKWS：具有可微架构搜索的关键字发现

智能音频设备由一个永远在线的轻量级关键字识别程序进行门控，以降低功耗。然而，设计具有高精度和低延迟的模型以实现准确和快速的响应是具有挑战性的。已经为开发端到端神经网络做出了许多努力，其中采用深度可分离卷积、时间卷积和 LSTM 作为构建单元。尽管如此，这些以人类专业知识设计的网络可能无法在广阔的搜索空间中实现最佳权衡。在本文中，我们建议利用可微神经架构搜索的最新进展来发现更有效的网络。我们发现的模型在 Google Speech Command Dataset v1 上达到了 97.2% 的 top-1 准确率。