PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
April 23, 2026Open Access

Deep Learning Approaches for Facial Emotion Recognition: A Comparative Review of Pretrained Models

View Full Paper
Ask AI
Bookmark
Share

Authors

GKGalina Kim

Discussion

Loading...

Member takes

Overview

This comparative review evaluates pretrained deep learning models for facial emotion recognition, suggesting optimal choices for different scenarios.

Key Points

  • The study aims to systematically evaluate the performance of pretrained deep learning models in facial emotion recognition.
  • Reviewed multiple pretrained deep learning architectures including CNN and transformer models.
  • Consolidated results from recent publications focusing on accuracy, complexity, and practicality.
  • Analyzed famous models like VGGNet, ResNet, EfficientNet, and ViT.
  • CNN-based models achieved 70–74% accuracy on the FER2013 dataset.
  • Modern architectures, including EfficientNet and hybrid transformer models, reached 85–90% accuracy.
  • A framework was provided for selecting suitable pretrained models based on application requirements.

Cite This Study

Galina Kim (2026) studied this question.

synapsesocial.com/papers/69e9ba6b85696592c86eca57https://doi.org/10.5281/zenodo.19675524
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Facial Emotion Recognition Using Deep Learning: A Comparative Study of CNN, ResNet-18 and MobileNetV22026
  2. 2Facial Emotion Recognition Using Transfer Learning in the Deep CNN2021 · 333 citations
  3. 3Facial Expression Recognition Using Pre-trained Architectures2024 · 3 citations
  4. 4A Comparative of Facial Emotion Recognition Performance Based on Cnn, Vggnet And Resnet2025
  5. 5FN-DeepCNN: Facial Expression Recognition Using Fine-Tuned Deep Convolutional Neural Network2026