Images -> Pretrained Model -> Embbedings (1000 features) -> Fit into kNN Classifier (1 image learnt -> 1 point with 1000 co-ordinates in the kNN model)
Python 3.10.6
To run, replace the file path of the images inside knn.py
python knn.py
Here are some example with different values of distances
representing how close the test images is, compared to the findings in the dataset: