Face-Keypoints-detection-with-CNN-and-GPU

This code detects faces and key points in an image, using CNN and Haar Cascade with GPU.

This project is based on the project I did for the Udacity Computer Vision Nanodegree. The objective is to detect faces in an image, detect their key points and print these key points in the image. To realize this project in Jupyter Notewook I used Python language, Pythorch to create the Neural Network and CV2 to work with the images.

The files are:

1. train-test-data.zip: This image file contains the images to train and test and 2 .csv with the name of the image and its Keypoints. it is not necessary to unzip the file in order to execute the code (to dowload this file: https://www.cs.tau.ac.il/~wolf/ytfaces/)
2. images folder: this folder contains the image of the ones I used at the end of the project.
3. models folder: folder with saved models
4. 'Face Key Points detection with CNN and GPU.ipynb': Jupyter notebook with all the code and its explanations.
5. detector_architectures folder: folder with the cascade codes.