face-clustering

AI model for grouping visually similar faces in images and video.

No input available.

Notes

Face Clusterer

Purpose

This model outputs the closest cluster ids that face embeddings belong to.

Architecture

The cluster model has been trained using face embeddings using the LOPQ approach.

Intended Use

Cluster model for face embeddings.

Limitations

The model requires retraining in case the face embedding changes significantly or test data distribution changes.

Taxonomy

The model is trained on face_msceleb1m.

  • ID
  • Model Type ID
    Clusterer
  • Input Type
    embeddings
  • Output Type
    clusters
  • Description
    AI model for grouping visually similar faces in images and video.
  • Last Updated
    Aug 24, 2022
  • Privacy
    PUBLIC
  • Toolkit
  • License
  • Share
  • Badge
    face-clustering