Face emotion recognition is crucial in modern AI applications, enabling systems to interpret human emotions from facial expressions. In this study, we develop a deep learning model to identify seven primary emotions: neutral, surprise, anger, disgust, fear, happiness, and sorrow. Our approach uses convolutional neural networks (CNNs) for feature extraction and applies preprocessing—such as face detection and alignment—to ensure reliable input data. Data augmentation improves generalization across varied facial appearances, and transfer learning enhances model efficiency by leveraging pre-trained weights. The proposed framework has applications in human-computer interaction, psychological analysis, and security systems. Challenges, including class imbalance and expression ambiguity, are addressed, and future work will explore multimodal integration (e.g., speech and text) for comprehensive emotion recognition. The model achieves an accuracy of 84.62%, surpassing traditional approaches. Preprocessing techniques such as face detection and data augmentation enhance robustness.