Person face re-identification using deep learning approach
N.V. Ravindhar, M. Mohamed Kasim Raja, S. Naveen Prabhu, H. Durgesh · 2023
This article discusses individual re-identification (Re-ID) using a deep learning approach. Its ability to identify important individuals from security camera footage has sparked a lot of attention. The rising amount of surveillance films has resulted in high computing and storage costs, which has posed a substantial hurdle for customers with limited resources. However, individual Re-ID over paid surveillance footage may pose a security risk by allowing the privacy of the uninvolved subject to be violated. As a result, the proposed system uses the haar cascade to successfully re-identify a person while safeguarding their privacy over contractual surveillance films. Both re-identification and privacy protection are possible with this technique for the person who was spotted. Deep learning will in particular extract elements that are useful for human re-identification and assess if a certain person&s;s face has been accurately identified.