search

kdmalc / fl-gestures

Develop a personalized sequential federated learning algorithm for real-time gesture classification
1 stars 0 forks source link

Test other embedding distances for clustering/matching #30

Open kdmalc opened 3 weeks ago

kdmalc commented 3 weeks ago

Cosine seems to work well for CLIP in the embedding space, don't see why that wouldn't work for us.

I already wrote code to compare the correlation, euclidean, and cosine distances, but only for the hierarchical code. Apply those same metrics to KMeans/KNN/etc.