Enhanced Facial Expression Recognition Based on ResNet50 with a Convolutional Block Attention Module
Zapisane w:
| Wydane w: | International Journal of Advanced Computer Science and Applications vol. 16, no. 1 (2025) |
|---|---|
| 1. autor: | |
| Wydane: |
Science and Information (SAI) Organization Limited
|
| Hasła przedmiotowe: | |
| Dostęp online: | Citation/Abstract Full Text - PDF |
| Etykiety: |
Nie ma etykietki, Dołącz pierwszą etykiete!
|
MARC
| LEADER | 00000nab a2200000uu 4500 | ||
|---|---|---|---|
| 001 | 3168740473 | ||
| 003 | UK-CbPIL | ||
| 022 | |a 2158-107X | ||
| 022 | |a 2156-5570 | ||
| 024 | 7 | |a 10.14569/IJACSA.2025.0160167 |2 doi | |
| 035 | |a 3168740473 | ||
| 045 | 2 | |b d20250101 |b d20251231 | |
| 100 | 1 | |a PDF | |
| 245 | 1 | |a Enhanced Facial Expression Recognition Based on ResNet50 with a Convolutional Block Attention Module | |
| 260 | |b Science and Information (SAI) Organization Limited |c 2025 | ||
| 513 | |a Journal Article | ||
| 520 | 3 | |a Deep learning techniques are becoming increasingly important in the field of facial expression recognition, especially for automatically extracting complex features and capturing spatial layers in images. However, previous studies have encountered challenges such as complex data sets, limited model generalization, and lack of comprehensive comparative analysis of feature extraction methods, especially those involving attention mechanisms and hyperparameter optimization. This study leverages data science methodologies to handle and analyze large, intricate datasets, while employing advanced computer vision algorithms to accurately detect and classify facial expressions, addressing these challenges by comprehensively evaluating FER tasks using three deep learning models (VGG19, ResNet50, and InceptionV3). The convolutional block attention module is introduced to enhance feature extraction, and the performance of the model is further improved by hyperparameter tuning. The experimental results show that the accuracy of VGG19 model is the highest 71.7\% before the module is integrated, and the accuracy of ResNet50 is the highest 72.4\% after the module is integrated. The performance of all models was significantly improved through the introduction of attention mechanisms and hyperparameter tuning, highlighting the synergistic potential of data science and computer vision in developing robust and efficient in facial expression recognition systems. | |
| 651 | 4 | |a Malaysia | |
| 653 | |a Feature extraction | ||
| 653 | |a Accuracy | ||
| 653 | |a Data analysis | ||
| 653 | |a Face recognition | ||
| 653 | |a Spatial data | ||
| 653 | |a Data science | ||
| 653 | |a Task complexity | ||
| 653 | |a Attention | ||
| 653 | |a Algorithms | ||
| 653 | |a Tuning | ||
| 653 | |a Computer vision | ||
| 653 | |a Modules | ||
| 653 | |a Deep learning | ||
| 653 | |a Machine learning | ||
| 653 | |a Datasets | ||
| 653 | |a Artificial intelligence | ||
| 653 | |a Neural networks | ||
| 653 | |a Emotions | ||
| 773 | 0 | |t International Journal of Advanced Computer Science and Applications |g vol. 16, no. 1 (2025) | |
| 786 | 0 | |d ProQuest |t Advanced Technologies & Aerospace Database | |
| 856 | 4 | 1 | |3 Citation/Abstract |u https://www.proquest.com/docview/3168740473/abstract/embedded/L8HZQI7Z43R0LA5T?source=fedsrch |
| 856 | 4 | 0 | |3 Full Text - PDF |u https://www.proquest.com/docview/3168740473/fulltextPDF/embedded/L8HZQI7Z43R0LA5T?source=fedsrch |