I am working on a project where I am trying to cluster different small image sets (each set has around 100 images) without knowing how many classes can each set contain (it can go from 2 to 4). I have tried using a convolutional autoencoder for feature extraction and dimensionality reduction and then ran HDBSCAN on the reduced images, but in vain. Does anyone have a suggestion for which feature extraction/clustering algorithm I can go with?
Asked
Active
Viewed 130 times
1
-
Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. – Community Apr 08 '22 at 18:22
-
how is the computer supposed to know the number of classes? theoretically it could put each input in its own class, or put all of them in a single class. – Christoph Rackwitz Apr 09 '22 at 09:20