Table 4 Classification accuracy of tiered-ImageNet dataset under different transduction settings.

From: Integrating deformable CNN and attention mechanism into multi-scale graph neural network for few-shot image classification

Model

Trans.

5-way 5-shot

Reptile

NO

66.47

Prototypical net

NO

69.67

MAML

NO

69.89

EGNN

NO

70.98

CA-EGNN

NO

72.31

MAML

BN

70.30

Reptile

BN

71.03

Relation Net

BN

71.31

MAML

YES

70.83

TPN

YES

72.58

EGNN

YES

80.15

CA-EGNN

YES

80.61

  1. Significant values are in bold.
Back to article page