Emotion recognition from facial images is an increasingly significant task across various domains. The objective of this paper is to tackle the emotion recognition classification problem using facial images. Deep Learning methodologies like Transfer learning and custom-trained Neural Networks, are employed to address this task. The classifier aims to distinguish the following emotions: happiness, neutrality, sadness, anger, surprise, disgust, and fear.
The dataset used to develop this project was sourced from Kaggle. Here are the dataset details:
- Source: Kaggle - FER2013 Dataset
- Size: 28,709 examples
- Image Type: 48x48 pixel grayscale images of human faces
- Classes: 7 emotion classes (Angry, Disgust, Fear, Happy, Sad, Surprise, Neutral)
- Total Size: 56.51 MB
This project was developed using the Google Colab environment for ease of sharing and collaboration. Jupyter notebooks are available on Google Colab and can be run directly in the cloud.
To run the notebooks on Google Colab, follow these steps:
- Open the desired notebook in the repository.
- Click on the Google Colab icon
at the top of the notebook.
- The notebook will open on Google Colab, ready for execution.
