Emotions and Gender Detection Using Artificial Intelligence

This report details the development and deployment of a real-time facial detection and affective state classification system, leveraging OpenCV and the DeepFace framework within the Google Colab environment. The system acquires live video input via a webcam interface, applies OpenCV’s Haar Cascade classifier for robust face localization, and utilizes pre-trained deep convolutional neural networks (CNNs) through DeepFace for multi-class emotion recognition. The processing pipeline encompasses image acquisition, grayscale conversion, normalization, facial region extraction, and emotion inference, with annotated visual feedback rendered directly on the video stream. Implementation challenges such as latency, model generalization, and environmental variability are analyzed, and prospective enhancements—such as model optimization, hardware acceleration, and integration of more advanced architectures—are proposed. Experimental results demonstrate the system’s viability for foundational facial affect analysis, with scope for improved precision and computational efficiency in future iterations.