Face Identification of Suspects Using Sequential -Deep Convolutional Neural Network

In terms of increasing terrorism, criminal behaviour, and anti-social events, there has been a need for safety systems to identify criminals. Face Recognition is one of the vibrant technologies that are very useful for person identification, especially in criminal investigation. Convolutional layers are used in our proposed system for feature extraction, and dense layers are used for classification. ReLU activation functions use four convolutional layers with 3 × 3 kernels and 128/64 filters to add non-linearity. Carefully placing dropout layers prevents overfitting, and Max Pooling layers down sample spatial dimensions. The model produces encouraging outcomes on a customized dataset comprising 4288 photos that capture different angles and concealed/hidden images of four subjects.