/Facial_Expression_DL

Using ResNet to Classify Facial Expression of Images.

Primary LanguageJupyter NotebookMIT LicenseMIT

Facial_Expression_DL

Using ResNet to Classify Facial Expression of Images.

🛠 Development Environment

  • Python 3.6.8
  • PyTorch 1.2.0

💾 Dataset

🎯 Accuracy

  • Overall Accuracy: 79.07%

    Test Accuracy of Surprise: 87.54%
    Test Accuracy of Fear: 44.59%
    Test Accuracy of Disgust: 40.00%
    Test Accuracy of Happiness: 89.54%
    Test Accuracy of Sadness: 73.85%
    Test Accuracy of Anger: 52.47%
    Test Accuracy of Neutral: 79.71%