Review Article

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

Table 5

Classification of driver distraction (attentive vs. distracted) for complete runs using LSTM networks: correctly classified runs, baseline runs correctly classified as “attentive”, runs with task correctly classified as “distracted” (in percent) for different numbers of memory blocks.

# Memory blocksCorrectlyAccuracyAccuracy
class. runs (%)baseline runs (%)runs with task (%)

70 88.4 62.6 98.0
90 90.9 72.3 97.8
100 92.9 84.5 96.0
11092.986.795.2
130 90.7 74.1 96.9
150 87.8 61.8 97.6