Image Segmentation by Edge Partitioning over a Nonsubmodular Markov Random Field

Table 1

Constant parameters are maintained for the Berkeley image set and the test methods. The top ranking results are written in bold, and the rankings are in parenthesis. For segmentation methods with a , a different segmentation number is assigned optimally for each image.

Method	PRI	VoI	BDE
MShift [1]	0.7958	1.9725	14.41
JSEG [20]	0.7756	2.3217	14.40 ()
GBIS [21]	0.7139	3.3949	16.67
NTP [22]	0.7521	2.4954	16.30
Saliency [23]	0.7758	1.8165	16.24
TBES [24]	0.80 ()	1.76 ()	—
CtoR [25]	0.806 ()	1.717 ()	11.57 ()
EPartition	0.804 ()	1.700 ()	11.16 ()
NCut [11]	0.772	2.259	12.94
SpecSeg [13]	0.8146	1.8545	12.21