MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITION
<p class="Abstract">Facial expression recognition (FER) on images with illumination variation and noises is a challenging problem in the computer vision field. We solve this using deep learning approaches that have been successfully applied in various fields, especially in uncontroll...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Institut Teknologi Sepuluh Nopember
2021-01-01
|
| Series: | JUTI: Jurnal Ilmiah Teknologi Informasi |
| Online Access: | http://juti.if.its.ac.id/index.php/juti/article/view/1031 |
| _version_ | 1818647148830392320 |
|---|---|
| author | Syavira Tiara Zulkarnain Nanik Suciati |
| author_facet | Syavira Tiara Zulkarnain Nanik Suciati |
| author_sort | Syavira Tiara Zulkarnain |
| collection | DOAJ |
| description | <p class="Abstract">Facial expression recognition (FER) on images with illumination variation and noises is a challenging problem in the computer vision field. We solve this using deep learning approaches that have been successfully applied in various fields, especially in uncontrolled input conditions. We apply a sequence of processes including face detection, normalization, augmentation, and texture representation, to develop FER based on Convolutional Neural Network (CNN). The combination of TanTriggs normalization technique and Adaptive Gaussian Transformation Method is used to reduce light variation. The number of images is augmented using a geometric augmentation technique to prevent overfitting due to lack of training data. We propose a representation of Modified Local Ternary Pattern (Modified LTP) texture image that is more discriminating and less sensitive to noise by combining the upper and lower parts of the original LTP using the logical AND operation followed by average calculation. The Modified LTP texture images are then used to train a CNN-based classification model. Experiments on the KDEF dataset show that the proposed approach provides a promising result with an accuracy of 81.15%.</p> |
| first_indexed | 2024-12-17T00:57:56Z |
| format | Article |
| id | doaj.art-4e1fb7e6ee53431a91a1cca7b187c530 |
| institution | Directory Open Access Journal |
| issn | 1412-6389 2406-8535 |
| language | English |
| last_indexed | 2024-12-17T00:57:56Z |
| publishDate | 2021-01-01 |
| publisher | Institut Teknologi Sepuluh Nopember |
| record_format | Article |
| series | JUTI: Jurnal Ilmiah Teknologi Informasi |
| spelling | doaj.art-4e1fb7e6ee53431a91a1cca7b187c5302022-12-21T22:09:33ZengInstitut Teknologi Sepuluh NopemberJUTI: Jurnal Ilmiah Teknologi Informasi1412-63892406-85352021-01-01191101810.12962/j24068535.v19i1.a1031491MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITIONSyavira Tiara Zulkarnain0Nanik Suciati1Institut Teknologi Sepuluh NopemberInstitut Teknologi Sepuluh Nopember<p class="Abstract">Facial expression recognition (FER) on images with illumination variation and noises is a challenging problem in the computer vision field. We solve this using deep learning approaches that have been successfully applied in various fields, especially in uncontrolled input conditions. We apply a sequence of processes including face detection, normalization, augmentation, and texture representation, to develop FER based on Convolutional Neural Network (CNN). The combination of TanTriggs normalization technique and Adaptive Gaussian Transformation Method is used to reduce light variation. The number of images is augmented using a geometric augmentation technique to prevent overfitting due to lack of training data. We propose a representation of Modified Local Ternary Pattern (Modified LTP) texture image that is more discriminating and less sensitive to noise by combining the upper and lower parts of the original LTP using the logical AND operation followed by average calculation. The Modified LTP texture images are then used to train a CNN-based classification model. Experiments on the KDEF dataset show that the proposed approach provides a promising result with an accuracy of 81.15%.</p>http://juti.if.its.ac.id/index.php/juti/article/view/1031 |
| spellingShingle | Syavira Tiara Zulkarnain Nanik Suciati MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITION JUTI: Jurnal Ilmiah Teknologi Informasi |
| title | MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITION |
| title_full | MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITION |
| title_fullStr | MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITION |
| title_full_unstemmed | MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITION |
| title_short | MODIFIED LOCAL TERNARY PATTERN WITH CONVOLUTIONAL NEURAL NETWORK FOR FACE EXPRESSION RECOGNITION |
| title_sort | modified local ternary pattern with convolutional neural network for face expression recognition |
| url | http://juti.if.its.ac.id/index.php/juti/article/view/1031 |
| work_keys_str_mv | AT syaviratiarazulkarnain modifiedlocalternarypatternwithconvolutionalneuralnetworkforfaceexpressionrecognition AT naniksuciati modifiedlocalternarypatternwithconvolutionalneuralnetworkforfaceexpressionrecognition |