Table 6 Effect of different modules (%)

From: OBI-CMF: Self-supervised learning with contrastive masked frequency modeling for oracle bone inscription recognition

HOG

CL

Top-1

Top-5

×

×

94.86

98.81

✓

×

94.97

98.81

×

✓

94.97

98.85

✓

✓

95.08

98.86

  1. Values in bold indicate optimal results.
  2. CL stands for contrastive learning. HOG stands for reconstructing HOG task.
Back to article page