search

DecodingRaphael / unraphael

Python toolkit for unraveling image similarity with a focus on artistic practice
https://unraphael.readthedocs.io
Apache License 2.0
2 stars 1 forks source link

In cluster process, the similarity index is wrongly calculated (always 1 or 0) #75

Closed ThijsVroegh closed 1 month ago

ThijsVroegh commented 2 months ago

As part of the cluster process, the similarity values are not correct (these are always 1 or 0). When not using the pipeline but directly inserting images in the get_image_similarity() function in image_clustering.py, the similarity values do appear correct.