The creation of an attendance system based on facial recognition is presented in this paper. Convolutional neural networks (CNNs) with the ResNet architecture are used to improve accuracy. Conventional attendance systems are prone to mistakes, proxy attendance, and inefficiencies since they frequently rely on manual entry or RFID cards. Attendance management with facial recognition is automated, secure, and non-intrusive. Utilizing CNNs for deep learning, the suggested system makes use of the Residual Neural Network (ResNet) model, which is well-known for its reliable performance in image identification applications. Because ResNet may use skip connections to counteract the vanishing gradient issue, deeper networks can be trained more successfully and with greater precision. Taking into consideration changes in lighting, orientation, and facial emotions, the model is trained on a wide range of facial picture datasets.