Yazar "Yenilmez, Fulya" seçeneğine göre listele

Listeleniyor 1 - 2 / 2
  • Küçük Resim
    Öğe
    Explanatory comparative study of AI models in face expression recognition
    (Türk-Alman Üniversitesi Fen Bilimler Enstitüsü, 2023) Yenilmez, Fulya
    This thesis provides an analysis and comparison of five widely used CNN architec tures in face expression recognition task. The objective is to evaluate the performance of different CNN architectures available in the Keras library for this challenging com puter vision task. The chosen CNN architectures for comparison include VGG19, InceptionV3, ResNet152V2, MobileNetV2, and EfficientNetV2B1. Two different kinds of facial datasets are used in this research. The first dataset is Fer2013, a com monly used dataset in this domain known for its unbalanced structure. The second dataset is the FACES dataset from the Max-Planck Institute, comprising posed im ages of individuals with a balanced structure. Both of these datasets contain labeled face expression images. Pre-processing steps, such as rotation, shift, resizing, and rescaling, are applied to these images. Since this is a comparative analysis study, the same transfer learning steps are applied to all models. The results from this training are evaluated using test accuracy, which is necessary for analyzing every aspect of the study. To ensure a fair comparison, the same transfer learning steps are applied to all models. The models are trained by dividing the dataset into three sets; training data set, validation data set, and test data set. The comparative study reveals that each CNN architecture exhibits different levels of performance in facial expression recognition tasks. The study provides significant knowledge about the strengths and weaknesses of each CNN architecture in face ex pression recognition. Overall, this comparison study provides clarity on the effectiveness of VGG19, In ceptionV3, ResNet152V2, MobileNetv2, and EfficientNetV2B1 architectures in rec ognizing facial expressions. It serves as a valuable resource for researchers in the fields of computer vision and expression analysis
  • [ X ]
    Öğe
    Facial Expression Recognition using Recent Convolutional Neural Network Models
    (Institute of Electrical and Electronics Engineers Inc., 2023) Yenilmez, Fulya; Yildiz, Canan; Ugur, Mukden
    In recent years, deep learning architectures have demonstrated remarkable achievements on diverse computer vision tasks, including the recognition of facial expressions. The development of precise and robust facial expression recognition (FER) models has the potential to improve a wide range of applications in various fields, such as human-computer interaction, emotion analysis, marketing, robotics, psychology, and health care.Within this domain, several Convolutional Neural Network (CNN) architectures have been employed to yield promising results. In this paper, we undertake a comparative analysis with the latest iterations of five widely adopted CNN models for the FER task: VGG19, InceptionV3, ResNet152V2, MobileNetV2, and EfficientNetV2B1. We also explore the effect of transfer learning, specifically the difference between pretraining on VGGFace versus ImageNet. Our primary goal is to comprehensively assess these architectures under identical conditions and provide clarity on their performance in comparison to each other and to previous works using various training strategies. Our assessments allow the conclusion that increasing the size and depth of these common "backbone"models leads to only minor improvements. Augmenting these models with additional architectural elements such as attention mechanisms seems much more promising in comparison. Our results further show that pretraining on a domain-specific dataset can lead to significant improvements, demonstrating the importance of a large FER related dataset. © 2023 IEEE.