Superpixel CRF, as proposed in Fulkerson et. al 2009.
BK = BLOCK_TEST_SEGCRF() Initializes the block with the default options.
BK = BLOCK_TEST_SEGCRF(BK) Executes the block with options and inputs BK.
Required inputs:
- db
The database.
- qseg
The quick shift segmentations.
- segloc
The unary potentials, in the form output by BLOCK_TEST_SEGLOC()
- traincrf
Parameters for the crf, from BLOCK_TRAIN_CRF()
Options:
- bk.rand_seed
Set the random seeds before proceeding. Default of [] does not change the seeds.
- bk.restrict
Restrict the possible solutions of the CRF to include only those which have co-occurred in the training data. Default 0.
Fetchable attributes:
- test
Classification result (images). Returns [class confidence] for required input: seg_id.
- segtest
Classification result (superpixels). Returns [class confidence] for required input: seg_id. class is a vector Nx1 where N is the number of superpixels. confidence is a matrix Nx1 cell array, where each cell is a Cx1 vector expressing the confidence in each of C classes.