Table 16 Improvement in runtime when using Improved GLTP on CK+ dataset
From: Improved gradient local ternary patterns for facial expression recognition
| Â | Training | Testing | ||
|---|---|---|---|---|
| Â | 6-class | 7-class | 6-class | 7-class |
(a) Per fold improvement (seconds) | Â | |||
CV | 7.07 | 25.94 | 0.19 | 0.97 |
LOO | 8.93 | 29.34 | 0.006 | 0.015 |
(b) Percentage improvement (%) | Â | |||
CV | 29.9 | 53.7 | 7.1 | 19.6 |
LOO | 32.6 | 54.1 | 16.2 | 29.4 |