Table 1 Performance comparison of trajectory planning models (UniAD and VAD) incorporating different phases of human attention
From: Human and algorithmic visual attention in driving tasks
Model | Phase | L2 (m) ↓ | Col. (%)↓ | ||||||
|---|---|---|---|---|---|---|---|---|---|
1s | 2 s | 3 s | Avg. | 1 s | 2 s | 3 s | Avg. | ||
UniAD | Baseline | 0.29 | 0.89 | 1.53 | 0.90 | 0.15 | 0.20 | 0.53 | 0.29 |
t0-t1 | 0.29 | 0.87 | 1.49 | 0.88 | 0.17 | 0.25 | 0.67 | 0.36 | |
t1-t2 | 0.26 | 0.83 | 1.36 | 0.82 | 0.13 | 0.17 | 0.54 | 0.27 | |
t2-treaction | 0.30 | 0.93 | 1.57 | 0.92 | 0.18 | 0.27 | 0.73 | 0.39 | |
t0-treaction | 0.28 | 0.86 | 1.44 | 0.86 | 0.15 | 0.24 | 0.63 | 0.34 | |
VAD | Baseline | 0.41 | 0.70 | 1.05 | 0.72 | 0.07 | 0.17 | 0.41 | 0.22 |
t0-t1 | 0.39 | 0.71 | 1.04 | 0.71 | 0.08 | 0.18 | 0.43 | 0.23 | |
t1-t2 | 0.33 | 0.57 | 0.96 | 0.62 | 0.08 | 0.13 | 0.35 | 0.19 | |
t2-treaction | 0.41 | 0.73 | 1.32 | 0.92 | 0.10 | 0.23 | 0.47 | 0.27 | |
t0-treaction | 0.33 | 0.60 | 1.06 | 0.66 | 0.09 | 0.15 | 0.37 | 0.20 | |