advanced clustering-based segmentation technique where mean shift moves towards dense regions in the data
- sliding window esque
- represent each pixel w feature vector
- define window over pixel
- identify all pixels within radius of feature vector
- compute mean of all pts inside window (center of gravity among neighbors of window)
- shift window towards mean feature location
- repeat until convergence (reaching high-density region)
- initialize multiple windows at random
- pixels that end up in same location belong to same cluster
- attraction basin: feature region for which all windows end up in same location
- speeds up
pro
- no need to specify K
- better at irregular cluster shapes than k-means
- does not assume prior shape
- single parameter
- robust to outliers
cons
- expensive - need to shift one window for every pixel - redundancy
- window size is hyperparameter that is hard to find
- does not scale well with dim of feature space