提出了跨域特征融合的端-边协同遮挡人脸识别方法，设计了端-边协作的人脸识别架构，充分利用终端和边缘的计算资源进行遮挡人脸的实时识别。提出了跨域人脸特征融合的方法，通过显式域人脸特征增强与隐式域人脸特征挖掘，进行人脸特征的融合提取。进而，提出时延优化的边缘识别任务调度方法，综合考虑边缘的任务负载、计算能力、带宽及时延容忍约束，动态调度人脸识别任务，在保证识别精度的同时最小化识别时延。实验结果表明，所提方法相比基线方法，在人脸识别任务精度基本持平的同时大幅提高了识别时延方面的性能。

The lack of facial features caused by wearing masks leads to the degradation of face recognition system performance. The methods of traditional occluded face recognition cannot integrate the computing resources of the edge layer and the device layer. Besides, previous research is short of comprehensive consideration of the facial characteristics including occluded and unoccluded parts. To solve the above problems, a device-edge collaborative occluded face recognition method based on cross-domain feature fusion is put forward. Specifically, the device-edge collaborative face recognition architecture gets the utmost out of device and edge resources for real-time occluded face recognition. Then a cross-domain facial feature fusion method is presented which combines both the explicit domain and the implicit domain facial. Furthermore, a delay-optimized edge recognition task scheduling method is devised, which comprehensively considers the task load, computing power, bandwidth, and delay tolerance constraints of the edge. This method can dynamically schedule face recognition tasks and minimize recognition delay while ensuring recognition accuracy. The experimental results show that the proposed method achieves an average gain of about 21% in recognition latency, while the accuracy of the face recognition task is the same compared to the baseline method.