search

cobanov / tasnif

A Python toolkit for image clustering using deep learning, PCA, and K-means, with support for GPU and CPU processing. Simplify your image analysis projects with advanced embeddings, dimensionality reduction, and automated visual categorization.
MIT License
31 stars 3 forks source link

added optinal pca and optinal iteration #4

Closed Stealeristaken closed 4 months ago

Stealeristaken commented 4 months ago

I have added the optional execution of the PCA process and the ability to specify the number of iterations as an integer.

cobanov commented 4 months ago

Actually, I think it makes more sense to open a pull request for each change, let's not change more than one thing at the same time.

The change you made for kmeans definitely looks very useful and I find it logical.

The vector length of the model output is really long, yes, you kept it optional, but in any case, I think it makes sense to keep PCA, running Kmeans without reducing the size does not make much sense to me.

Stealeristaken commented 4 months ago

I opened this pr all at once because the changes were small imo. x)

Yes, you are right. In many cases, it doesn't make sense to not perform PCA. But, as you said, I kept it optional by setting the default to True. However, you are the boss, so if you want me to change it, please create a comment on the code line, and I will do it ASAP

cobanov commented 4 months ago

No worries! It's good, thank you for your contribution!

cobanov commented 4 months ago

https://github.com/cobanov/tasnif/releases/tag/0.1.10