A Face Detection and Recognition system based on Eigenface method
-
The file "Face_Detection_Recognition" contains the complete set of codes which were used for building the final functional code: "PCA-Face". It contains the hand-coded versions of the neural network(s), face alignment code(though, the method applied did not work on the given dataset) and other experiments performed with proper references (giving credit to original authors).
-
Pre-trained models have been saved in the Output-files folder (file names beginning with "CLASS").
-
Report present in this folder contains complete description of all the experiments conducted and the resources used, steps in pipeline which are to be there in an ideal face recognition system, and how much of it have been implemented. It also contains a pseudo code of the submitted code.
If running the code "PCA-Face.ipynb" on Google Colab:
- Upload the "PCA-Face.zip" file.
- Continue execution
The "PCA-Face.zip" file contains the necessary training and testing samples in the folder PCA-Face. An empty folder named "Output-images" has been included in the zip file, into which the code stores the output files, such as plots, images, trained model files, etc., If one is interested in knowing more about the neural networks and other things tried before the final code (PCA_Face.ipynb) was made, kindly use the Face_Detection_and_Recognition.ipynb file.