Skip to main content

Table 5 Ablative study on sub-JHMDB dataset (%)

From: Human activity prediction using saliency-aware motion enhancement and weighted LSTM network

Settings

SME

Weighted module

Accuracy with half videos

Accuracy with full videos

Setting 1

x

x

48.9

59.2

Setting 2

x

✓

59.3

63.4

Setting 3

✓

x

61.5

69.7

Setting 4

✓

✓

66.8

78.1