Deep Learning Vs. Traditional Machine Learning For Facial Emotion Recognition In Psychiatric Applications
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The early diagnosis of mental illnesses including stress, worry, and depression may be greatly aided by facial expression identification. In order to classify six fundamental emotions (anger, disgust, fear, happiness, sadness, and surprise), this study assesses the performance of five machine learning models: K-Nearest Neighbours (KNN), Support Vector Machine (SVM), Random Forest (RF), Multilayer Perceptron (MLP), and a general Neural Network (NN). Using the PyTorch and scikit-learn frameworks, experiments were carried out on a system with an Intel i7-12700F CPU, 32 GB RAM, and an RTX 3070 GPU. Accuracy, precision, recall, and F1-score were used to compare models. With an overall accuracy of 95% and high F1-scores across all emotion classes, the results show that the MLP model performs better than all other approaches, proving its capacity to capture intricate, non-linear emotional patterns. While a traditional NN performed worse (81%), KNN, RF, and SVM obtained modest accuracies (62–84%). While simpler models like KNN are still competitive for small datasets, deep neural networks perform better for larger or more homogeneous data, according to analysis across several datasets. The results encourage the inclusion of AI-driven technologies into real-time mental health monitoring and intervention systems by highlighting the efficacy of deep learning for accurate emotion recognition.
Fields of Study & Subjects
Emotion and Mood Recognition
Face and Expression Recognition
Face recognition and analysis
Publication details
- Journal / source
- International Journal of Advances in Signal and Image Sciences
- Volume
- 12
- Issue
- 3s
- Pages
- 1154–1174
- Publication date
- 9 March 2026
- Language
- EN