Smart Doorbell with Emotion Detection

This paper presents a smart doorbell system that integrates computer vision and deep learning for real-time facial emotion recognition. Utilizing the ESP32-CAM module and a convolutional neural network (CNN) trained on the FER-2013 dataset, the system captures and classifies visitor emotions upon doorbell press. If negative emotions such as anger, fear, or sadness are detected, the system activates a buzzer and sends an email alert to the homeowner. Experimental results demonstrate an average emotion classification accuracy of 82%, highlighting the model's robustness across diverse lighting and facial orientations. This system offers a proactive and cost-effective solution for enhancing home security using AI.