tattle-made / data-experiments

EDA, small data science projects and other experiments

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Include similar images in weekly image clustering

variablenerd opened this issue · comments

commented

The weekly clustering notebook at https://github.com/tattle-made/data-experiments/blob/master/clustering-weekly.ipynb clusters images with identical image hashes. Expanding these clusters to include image hashes within a certain Hamming distance threshold would be useful for creating clusters of similar (and not just identical) images. This has been explored in https://github.com/tattle-made/data-experiments/blob/master/clustering-phash.ipynb.

Task : Modify the weekly image clustering script to include similar images.