Unrestricted Face Recognition Algorithm Based on Transfer Learning on Self-Pickup Cabinet

In the contactless delivery scenario, self-pickup cabinet is an important terminal device, and face recognition one of efficient ways to achieve access express delivery. order effectively recognize images under unrestricted environments, algorithm based on transfer learning proposed in this study. First, region extraction network faster RCNN improved improve speed algorithm. Then, first applied between large ImageNet dataset image restricted conditions. The second conditions datasets. Finally, processed by enhancement increase its similarity with image, so that can be carried out effectively. Experimental results show has better rate CASIA-WebFace dataset, FLW MegaFace dataset.