Skip to content

Clustering

Finding similar media

media-similarity can take several media, tag them and combine them into clusters.

media-similarity cluster image1.png image2.png image3.png \
  --media-type IMAGE \
  --tagger gemini

from media_similarity import MediaClusteringRequest, MediaSimilarityService

service = MediaSimilarityService()

request = MediaClusteringRequest(
  media_paths=[
    'image1.png',
    'image2.png',
    'image3.png',
    ],
    media_type='IMAGE',
    tagger_type='gemini',
)

clusters = service.cluster_media(request)

Working with clusters

Once cluster_media is executed you get a ClusteringResults object which contains:

  • clusters - mapping between media url and its assigned cluster
  • adaptive_threshold - minimal value for defining similar media.
  • graph -mapping with nodes and edges.

You can get clusters two ways:

  • via .clusters property as a dict
  • via .to_garf_report method as a GarfReport which contains two columns - media_url and cluster_id.