Scholarlink

In this study, we explore various Deep Learning architectures for facial expression recognition, with a focus on Transfer Learning. This approach leverages features extracted from large-scale datasets, such as the ImageNet dataset, to enhance the training of models on smaller datasets like the CK+ dataset. By utilizing pre-trained models such as MobileNet-V3 and VGG19, we aim to retain the generalization capacity while improving the performance of facial expression recognition on more specific datasets. The results demonstrate that MobileNet-V3 and VGG19 models achieve superior accuracy compared to traditional sequential models, confirming their effectiveness in facial expression recognition tasks.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Transfer Learning Approach for Facial Emotion Recognition Using Deep Convolutional Neural Network

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Fouad Lehlou,
  • Abdelaali Kemmou,
  • Jalal Laassiri,
  • Adil El Makrani

摘要

In this study, we explore various Deep Learning architectures for facial expression recognition, with a focus on Transfer Learning. This approach leverages features extracted from large-scale datasets, such as the ImageNet dataset, to enhance the training of models on smaller datasets like the CK+ dataset. By utilizing pre-trained models such as MobileNet-V3 and VGG19, we aim to retain the generalization capacity while improving the performance of facial expression recognition on more specific datasets. The results demonstrate that MobileNet-V3 and VGG19 models achieve superior accuracy compared to traditional sequential models, confirming their effectiveness in facial expression recognition tasks.