Face Clustering using Python

PROBLEM STATEMENT

We have given a dataset of face images and we have to separate similar face images into their respective directories.

Usually dataset of images contains thousands of images and labeling them manually can be a time consuming task.

Dataset

We have a dataset of unlabeled Face images extracted from video file using opencv.

Algorithm

K-means Clustering (Unsupervised Learning)

Code Snippets

So below are the code snippets for the whole python program.

Result

So after after running the above code a new directory will be created called "cluster".This directory contains the respective label's directory (label0 and label1).

So below are the final images separated in their respective label directories .

Conclusion

So this is how we can reduce the manual task of separating similar images into groups and label them respectively.