Enhanced Emotion Recognition on the FER-2013 Dataset by Training VGG from Scratch
Journal Title: International Journal of Innovations in Science and Technology - Year 2024, Vol 6, Issue 4
Abstract
Recognizing facial emotions is still a major obstacle in computer vision, particularly when dealing with complex datasets such as FER-2013. Advancements in deep learning have simplified the process of achieving high accuracy, yet obtaining high accuracy on the FER-2013 dataset with traditional methods remains challenging. The aim of this research is to analyze the effectiveness of Convolutional Neural Networks (CNNs) utilizing VGG16 and ResNet50 architectures through three different training methods: training from scratch, with transfer learning, and fine-tuning. Our research demonstrates that while training VGG16 from scratch achieved a validation accuracy of 67.23%, fine-tuning produced a slight reduction in performance at 64.80%. Conversely, ResNet50 struggled across all approaches, with the highest validation accuracy being only 54.69% when trained from scratch. We offer an in-depth analysis of these methodologies by utilizing confusion matrices, training durations, and accuracy measures to showcase the balance between computational expenses and model effectiveness. Our results indicate that, although transfer learning and fine-tuning offer rapid convergence, training from scratch may still be necessary for specialized feature learning in complex FER tasks. These results help in the continuous work of improving emotion recognition systems by maintaining a balance between accuracy and computational efficiency.
Authors and Affiliations
Muhammad Talha Jahangir, Syed Yameen Ali, Ali Hasnain, Muhammad Faizan, Mohsin Jabbar, Muhammad Kashan Basit, Abdul Rafay
Keywords
Related Articles
- EP ID EP760548
- DOI -
- Views 29
- Downloads 0
Barriers to AI Adoption in Education: Insights from Teacher's Perspectives
Artificial intelligence in education is capable of offering significant benefits in the form of content generation, personalized learning, assistance in administration, and analytical reports. Despite the benefits, the...
Early Detection and Classification of Lung Cancer using Segment Anything Model 2 and Dense Net
Lung cancer is one of the most perilous diseases worldwide with high incidence and low survival rates due to late diagnosis. Accurate detection and diagnosis of lung nodules is important for early-stage detection. Mach...
Exploring the Efficacy of CNN Architectures for Esophageal Cancer Classification Using Cell Vizio Images
Esophageal cancer, as with the global burden of disease, is usually due to Barrret's esophagus and gastroesophageal reflux disease. Fortunately, the disease is amenable to early detection; however, early diagnosis has...
Unlocking Potential: Personality-Aware TVET Course Recommendations Revolutionize Skill Development
Personality is a complex amalgamation of ideas, behaviors, and social constructs that shape our self-perception and influence our interactions with others. It tends to remain relatively stable over time. The developmen...
Mobile Legends Win Rate Prediction and Team Recommendation Using Switched Hero Roles
Introduction: Mobile Legends Bang Bang (MLBB) falls under the category of a multi-line battle arena game which requires players to have strong skills and strategic gameplay; team composition is an important factor infl...