Effective and Generalizable Graph-Based Clustering for Faces in the Wild

Table 4

Comparison of clustering results in the AR Face database.

Method	F-measure	FBcubed	Clusters
ResNet-29 + k-means ()	0.199	0.245	50
ResNet-29 + k-means ()	0.362	0.383	153
ResNet-29 + k-means ()	0.322	0.348	239
ResNet-29 + GLC	0.392	0.419	309
Approximate Rank-order [7]	—	—	—
ConPaC [4]	—	—	—
ResNet-29 + Approximate Rank-order	0.388	0.436	239
ResNet-29 + ours (proposed)	0.447	0.498	153

The true number of identities is 50, and the total number of face images is 2,600.