Review Article

Emotion on the Road—Necessity, Acceptance, and Feasibility of Affective Computing in the Car

Table 4

Sample-wise classification of driver distraction (attentive vs. distracted) using LSTM networks: accuracy, unweighted recall, unweighted precision, and F1-measure for different numbers of memory blocks.

# Memory blocksAccuracy (%)Recall (%)Precision (%) (%)

70 87.2 78.8 85.6 82.1
90 89.0 82.6 87.0 84.8
100 91.1 88.7 88.0 88.4
11091.389.587.988.7
130 89.0 83.3 86.4 84.9
150 86.2 77.2 84.2 80.6