Back to projects
Facial Emotion Detection
Abstract.Classifying emotions such as sadness, joy and surprise from facial expressions.
- Role
- Developer
- Period
- Mar to May 2024
- Context
- Personal project
- Team
- Solo
| ~16k | Face images |
| 4 | Models compared |
01How it works
- Input Capture image A face image from the camera or the dataset.
- Step 02 Detect the face OpenCV finds the face, converts to grayscale and preprocesses.
- Step 03 Extract features Expression features for classification.
- Output Classify & evaluate Four models compared with confusion matrices and heatmaps.
02What I built
- Collected and preprocessed a dataset of about 16,000 face images with OpenCV (face detection, grayscale conversion).
- Compared KNN, logistic regression, naive Bayes and random forest classifiers.
- Evaluated with confusion matrices and heatmaps, and improved accuracy through feature work.
03Technologies
- Python
- OpenCV
- scikit-learn
- NumPy
- Matplotlib
- Seaborn