This brief presents a low-power attention-based automatic speech recognition (ASR) processor achieving real-time recognition capability. The proposed attention window algorithm, compact end-to-end neural-network topology, and efficient computation dataflow effectively minimize the hardware complexity and power consumption, enabling a fully integrated low-power ASR processor solution without the necessity of any off-chip memory resource. The proposed design techniques reduced 98.9% weight memory and 92.1% power consumption with minimal degradation of 2.24% in recognition accuracy. The proposed ASR processor operates at 100MHz with 1.7mW at 0.9V, demonstrating 2x and 1.68x performance improvements in speed and power, respectively, compared to the previous ASR designs that require additional supports of off-chip memory or external decoder.

中文翻译：

---

一个完全集成的 1.7mW 基于注意力的自动语音识别处理器

本简介介绍了一种基于注意力的低功耗自动语音识别 (ASR) 处理器，可实现实时识别能力。所提出的注意力窗口算法、紧凑的端到端神经网络拓扑和高效的计算数据流有效地降低了硬件复杂度和功耗，实现了完全集成的低功耗 ASR 处理器解决方案，无需任何片外内存资源. 所提出的设计技术减少了 98.9% 的权重记忆和 92.1% 的功耗，同时识别准确度降低了 2.24%。与需要额外支持片外存储器或外部解码器的先前 ASR 设计相比，所提议的 ASR 处理器在 100MHz 下以 1.7mW 在 0.9V 下运行，在速度和功率方面分别表现出 2 倍和 1.68 倍的性能改进。