Table 2.
Predictability for the naive and semantic GHMM models with and without the grasping state feature.
| Naive model | With GS (%) | Without GS (%) | Semantic model | With GS (%) | Without GS (%) |
|---|---|---|---|---|---|
| Lined-up | 77.1 | 98.5 | Lined-up | 70.5 | 98.5 |
| Triangular | 78.2 | 99.0 | Triangular | 71.9 | 99.2 |