In this project, we studied, implemented and evaluated the methods described in this paper for Face Recognition, particulary FisherFaces and its comparison with Eigenfaces. Fisherface method tackles various challenges faced in face recognitoin such as illumination variations or facial expressions. Used Yale, YaleB and CMU face image datasets for evaluation. Also, implemented Glass recognition on Yale dataset.

All the code is present in src directory and some notebooks for plotting results and analyzing Yale dataset are in notebooks folder. For running the code, go inside src and run

python3 main.py --dataset <dataset_name>

where <dataset_name> $\in$ [yale, yaleB, cmu]

It will run the 3 methods (eigenfaces, eigenfaces leaving top 3, fisherfaces) on the dataset provided as arguement and output the evaluation results. For yale dataset, it will also perform Glass recognition and display its results using leaving one out method

• Atishay Jain (210050026)
• Cheshta Damor (210050040)
• Kanad Shende (210050078)