Citation

Hameed, Nouman and Khan, Adil and Zhou, Zhiyuan and Roslee, Mardeni and Ullah, Yasir and An, Shuo (2025) Deep Learning for Emotion Understanding: An Improved VGG16-Based Approach to Expression Recognition. In: 2025 Multimedia University Engineering Conference (MECON), 21-23 July 2025, Cyberjaya, Malaysia.

Text IEEE Xplore Full-Text PDF_45.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

Artificial intelligence and computer vision focus mainly on facial expression recognition as a fundamental technique to comprehend human emotions. A CNN-based facial expression recognition method is proposed to use VGG's 16 excellent feature extraction methods for analyzing emotion in faces. The VGG architecture uses its deep structure, including successive convolutional, pooling, and fully connected layers, to automatically find specific facial expression characteristics in images. A ReLU activation function is used to train the model against 35,886 grayscale images of seven basic expressions (e.g., happy, sad, angry, fear, disgust, surprise, and neutral) present in the FER2013 dataset with optimized hyperparameters. Experimental results demonstrate 70% accuracy on the test set, with extremely good performance in recognizing the expression neutral (79%), and happy (87%). Experimental findings show the efficacy of the model in detecting diverse facial expressions. The visualization techniques validate how VGG efficiently detects essential facial expression patterns. The research provides beneficial theoretical aspects of deep learning techniques for emotion comprehension while demonstrating practical implementations in human-machine interaction systems.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Convolutional Neural Network
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5101-6720 Telecommunication. Including telegraphy, telephone, radio, radar, television
Divisions:	Faculty of Artificial Intelligence & Engineering (FAIE)
Depositing User:	Ms Rosnani Abd Wahab
Date Deposited:	18 Mar 2026 04:30
Last Modified:	18 Mar 2026 05:18
URII:	http://shdl.mmu.edu.my/id/eprint/15536

Downloads

Edit (login required)