🧠 Facial Emotion Recognition — Comparative Analysis

📌 Overview

Comparative study of CNN and SVM models for facial emotion recognition across two real-world datasets (CK+ and RAF-DB). Includes full data preprocessing pipeline, model training, and performance evaluation. Research published in Springer, 2024.

📦 Datasets Used

• CK+ — Extended Cohn-Kanade Dataset
• RAF-DB — Real-world Affective Faces Database

🛠️ Tech Stack

Python | TensorFlow/Keras | Scikit-learn | Pandas | NumPy | Matplotlib

📊 Key Results

Dataset	Model	Accuracy
CK+	CNN	96%
CK+	SVM	97%
RAF-DB	CNN	85%
RAF-DB	SVM	77%

CNN outperformed SVM on the more complex RAF-DB dataset, demonstrating superior ability to handle varied lighting and facial orientations.

🔄 Pipeline Stages

1. Data Loading & Extraction
2. Image Preprocessing & Normalization
3. Feature Engineering
4. Model Training (CNN & SVM)
5. Performance Evaluation & Comparison

📄 Publication

Gobind Kumar et al., "Facial Emotion Recognition: Review and Perspectives" Springer, 2024 https://link.springer.com/chapter/10.1007/978-981-97-8526-1_43

👤 Author

Gobind Kumar | gobind1721@gmail.com