Table 2 Prediction accuracy comparison of different modalities on UT-Interaction set 2 (%)
From: Human activity prediction using saliency-aware motion enhancement and weighted LSTM network
Modality | Accuracy with half videos | Accuracy with full videos |
|---|---|---|
RGB-SME | 84.7 | 91.5 |
OF-SME | 75.2 | 92.6 |
Two-modality fusion | 90.2 | 95.1 |