Research Article

Multilabel Image Annotation Based on Double-Layer PLSA Model

Table 2

The results of contrast experiments: AR (average recall) and AP (average precision).

ā€‰ 49 labels 260 labels
ā€‰ AR AP AR AP

TM 0.34 0.20 0.04 0.06
CMRM 0.48 0.40 0.09 0.10
CRM 0.70 0.59 0.19 0.16
PLSA-WORDS 0.71 0.56 0.20 0.14
PLSA-FUSION 0.76 0.65 0.22 0.19
Double-layer PLSA 0.74 0.70 0.25 0.20