Table 1.
This table shows the Laban error distribution.
| Recognized emotion | |||||
|---|---|---|---|---|---|
| Intended emotion | Happy | Sadness | Fear | Anger | Neutral |
| Happy | 81.39 | 1.36 | 1.65 | 2.07 | 13.52 |
| Sadness | 0.61 | 78.57 | 5.07 | 5.38 | 10.37 |
| Fear | 0.61 | 23.12 | 51.15 | 9.14 | 15.98 |
| Anger | 2.57 | 3.67 | 26.17 | 47.29 | 20.31 |
| Neutral | 11.77 | 11.45 | 5.00 | 4.35 | 67.42 |
All numbers represent percentages. Rows represent the intended (“correct”) emotions and columns represent the emotion that was recognized. Thus, the number at each cell represents the percentage of clips recognized as the column emotion from all clips associated with the row emotion. For example: 47.29% of all anger associated clips were recognized as anger and 26.17% of all anger clips were recognized as fear.