https://github.com/andreaschatzopoulos/face-landmark-detector
Facial landmark detection using HOG features and Ridge Regression. Simple, effective, and fast – no deep learning required.
https://github.com/andreaschatzopoulos/face-landmark-detector
computer-vision face-detection hog image-processing landmark-detection python ridge-regression scikit-learn
Last synced: about 1 month ago
JSON representation
Facial landmark detection using HOG features and Ridge Regression. Simple, effective, and fast – no deep learning required.
- Host: GitHub
- URL: https://github.com/andreaschatzopoulos/face-landmark-detector
- Owner: AndreasChatzopoulos
- Created: 2025-06-17T09:55:46.000Z (12 months ago)
- Default Branch: main
- Last Pushed: 2025-06-17T10:01:43.000Z (12 months ago)
- Last Synced: 2025-06-17T11:19:41.987Z (12 months ago)
- Topics: computer-vision, face-detection, hog, image-processing, landmark-detection, python, ridge-regression, scikit-learn
- Homepage:
- Size: 1.95 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Facial Landmark Detection with HOG and Ridge Regression
This project detects 5 facial landmarks using machine learning with handcrafted image features.
## Features
- Data augmentation (flipping, brightness/contrast)
- Feature extraction using Histogram of Oriented Gradients (HOG)
- Ridge Regression model for landmark coordinate prediction
- Comparative baseline using SIFT + PCA
## Dataset
- 2811 training images and 554 test images (96x96 grayscale)
## Results
- HOG + Ridge Regression achieved 4.11 px mean landmark error
- Outperformed SIFT baseline
## How to Run
Open the `task_2_final.ipynb` notebook and execute all cells. Ensure the dataset is correctly referenced and dependencies installed.
## Requirements
```bash
pip install numpy pandas matplotlib scikit-image scikit-learn opencv-python
```
## Author
[Your Name]