Face Clustering by Karan Singh

A custom made face clustering algorithm which uses dlib's face recognition model to get face embeddings and then perform clustering on it efficiently and accurately with changable parameters

Installing Libraries

To successfully run the code first install the required libraries using

  pip install cmake
  pip install -r requirements.txt

Inference

To run the inference use the Karan_face_cluster.py file with changable parameters:

• faces_dir path to directory consisting faces
• face_distance_tolerance (default 0.5) Tolerance value used to compare two faces
• min_faces_cluster (default 15), Minimum faces to form cluster
• percentage_for_non_identified (default 0.35) Percentage value for comparing non identified faces to already made clusters

Note: This algorithm works only on datasets comparing already cropped faces, not full images

  python Karan_face_cluster.py --faces_dir 'test_250' --min_faces_cluster 20

Authors

• @Karan Singh