def spectral_clustering_sb(
affinity,
n_clusters=8,
n_components=None,
random_state=None,
n_init=10,
):
"""Performs spectral clustering.
Arguments
---------
affinity : matrix
Affinity matrix.
n_clusters : int
Number of clusters for kmeans.
n_components : int
Number of components to retain while estimating spectral embeddings.
random_state : int
A pseudo random number generator used by kmeans.
n_init : int
Number of time the k-means algorithm will be run with different centroid seeds.
Returns
-------
labels : array
Cluster label for each sample.
Example
-------
>>> import numpy as np
>>> from speechbrain.processing import diarization as diar
>>> affinity = np.array([[1, 1, 1, 0.5, 0, 0, 0, 0, 0, 0.5],
... [1, 1, 1, 0, 0, 0, 0, 0, 0, 0],
... [1, 1, 1, 0, 0, 0, 0, 0, 0, 0],
... [0.5, 0, 0, 1, 1, 1, 0, 0, 0, 0],
... [0, 0, 0, 1, 1, 1, 0, 0, 0, 0],
... [0, 0, 0, 1, 1, 1, 0, 0, 0, 0],
... [0, 0, 0, 0, 0, 0, 1, 1, 1, 1],
... [0, 0, 0, 0, 0, 0, 1, 1, 1, 1],
... [0, 0, 0, 0, 0, 0, 1, 1, 1, 1],
... [0.5, 0, 0, 0, 0, 0, 1, 1, 1, 1]])
>>> labs = diar.spectral_clustering_sb(affinity, 3)
>>> # print (labs) # [2 2 2 1 1 1 0 0 0 0]
"""
random_state = _check_random_state(random_state)
n_components = n_clusters if n_components is None else n_components
maps = spectral_embedding_sb(
affinity,
n_components=n_components,
drop_first=False,
)
_, labels, _ = k_means(maps,
n_clusters,
random_state=random_state,
n_init=n_init)
return labels
def clusterSpectralEmbeddings(self, affinity, n_init=10, cuda=False):
spectral_emb = self.getSpectralEmbeddings(affinity,
n_spks=self.n_clusters,
drop_first=False,
cuda=cuda)
_, labels, _ = k_means(spectral_emb,
self.n_clusters,
random_state=self.random_state,
n_init=n_init)
return labels
def cluster_embs(self, emb, k):
"""Clusters the embeddings using kmeans.
Arguments
---------
emb : array (n_samples, n_components)
Spectral embedding for each sample with n Eigen components.
k : int
Number of clusters to kmeans.
Returns
-------
self.labels_ : self
Labels for each sample embedding.
"""
_, self.labels_, _ = k_means(emb, k)
