Table 5 Analysis of different layers on the results of the three data sets. Significant values are in [bold].
From: Transductive zero-shot learning via knowledge graph and graph convolutional networks
Layers | AWA2 | ImageNet50 | ImageNet100 | ||||||
|---|---|---|---|---|---|---|---|---|---|
TGCNZ_1 | TGCNZ_2 | TGCNZ_3 | TGCNZ_1 | TGCNZ_2 | TGCNZ_3 | TGCNZ_1 | TGCNZ_2 | TGCNZ_3 | |
1 | 36.58 | 37.07 | 36.15 | 25.18 | 33.51 | 31.85 | 18.30 | 18.80 | 18.12 |
2 | 32.96 | 37.47 | 32.49 | 21.23 | 25.62 | 25.25 | 14.45 | 15.96 | 16.01 |
3 | 35.04 | 38.60 | 41.65 | 21.60 | 23.52 | 27.53 | 14.62 | 13.77 | 15.32 |
