研究人员开发了一种适用于可穿戴设备的新型深度学习架构，可显著降低心律失常检测的功耗。通过采用数据精度降低和近似乘法等技术，所提出的架构与现有模型相比，功耗降低了64.9%，同时保持了93.7%的高分类准确率。这项进展有望在不影响诊断性能的情况下延长可穿戴健康监测设备的电池寿命。 AI
影响 通过优化低功耗环境下的深度学习模型，延长可穿戴健康设备的电池寿命。
排序理由 该集群包含一篇详细介绍新方法和实验结果的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2607.14747v1 Announce Type: cross Abstract: Cardiovascular diseases are the leading cause of death worldwide, and conditions such as arrhythmia often require long-term monitoring for effective detection and diagnosis. However, current wearable monitoring devices are bulky, …
Cardiovascular diseases are the leading cause of death worldwide, and conditions such as arrhythmia often require long-term monitoring for effective detection and diagnosis. However, current wearable monitoring devices are bulky, uncomfortable, and typically rely on clinicians to…