Table 4.

Classification performance comparison to early works on the SenseEmotion Database in a LOSO cross-validation setting for the classification task $T_{0}vs.T_3$.

Approach	Description	Performance
Kalischek et al. [38]	Transfer Learning	$60.10\pm 00.06$
Thiam et al. [15]	Standardised Geometric Features	$\mathbf{66.22}\pm \mathbf{14.48}$
Our Approach	Motion Histogram Images	$60.86\pm 09.81$
Our Approach	Optical Flow Images	$62.70\pm 09.24$
Our Approach	Weighted Score Aggregation	$64.35\pm 10.40$

The performance metric consists of the average accuracy (in %) over the LOSO cross-validation evaluation. The best performing approach is depicted in bold and the second best approach is underlined.