Cost-Effective Person Identity Recognition based on Plantar Pressure Images

With the aging of world population structure, it is an important issue to identify patients in the long-term care environment. In order to keep privacy, this paper studies plantar pressure images as biometric features. However, most high-accuracy deep learning models require high computation resources. This paper proposed one cost-effective person identity recognition model. According to the initial experimental results, the proposed model has lower FLOPs and higher accuracy.