Table 14 Precision, recall, and F1-score comparison for anomaly detection on UCSD Ped1 dataset.

From: Design of an integrated model with temporal graph attention and transformer-augmented RNNs for enhanced anomaly detection

Model

Precision (%)

Recall (%)

F1-Score (%)

Proposed model (TGAT + TARNN)

92.1

89.4

90.7

Method4

84.5

82.7

83.6

Method8

81.2

86.5

83.7

Method15

78.9

80.3

79.6

Back to article page