ADE-8/12: 0.29FDE-8/12: 0.48
Papers with Code · CC BY-SA 4.0
Trajectory Prediction
Trajectory Prediction is the problem of predicting the short-term (1-3 seconds) and long-term (3-5 seconds) spatial coordinates of various road-agents such as cars, buses, pedestrians, rickshaws, and animals, etc. These road-agents have different dynamic behaviors that may correspond to aggressive or conservative driving styles.
<span class="description-source">Source: Forecasting Trajectory and Behavior of Road-Agents Using Spectral Clustering in Graph-LSTMs </span>
논문
ADE-8/12: 0.32FDE-8/12: 0.33
ADE-8/12: 0.20FDE-8/12: 0.31
ADE-8/12: 0.21FDE-8/12: 0.35
ADE-8/12: 0.23FDE-8/12: 0.39
ADE-8/12: 0.49
minADE (K=6): 0.79minFDE (K=6): 1.16MR (K=6): 0.12brier-minFDE (K=6): 1.75
ADE-8/12 @K = 20: 6.77FDE-8/12 @K= 20: 10.46
ADE-8/12: 0.21
ADE-8/12 @K = 20: 7.85FDE-8/12 @K= 20: 11.85
ADE-8/12 @K = 20: 6.52FDE-8/12 @K= 20: 10.61
mADEK @4.8s: 14.36mF DEK @4.8s: 25.99
ADE-8/12: 0.46
ADE-8/12 @K = 20: 10.27FDE-8/12 @K= 20: 19.71
ADE-8/12 @K = 20: 8.11FDE-8/12 @K= 20: 13.06ADE (8/12) @K=5: 11.64FDE(8/12) @K=5: 21.12
ADE-8/12: 0.19FDE-8/12: 0.40
ADE-8/12: 0.48FDE-8/12: 0.84
ADE-8/12: 0.33FDE-8/12: 0.33
ADE-8/12: 0.49
ADE-8/12: 0.54
BLEU: 29.7DTW R@1: 29.3Accuracy: 28.7
ADE-8/12: 0.37FDE-8/12: 0.71
ADE-8/12 @K = 20: 12.58FDE-8/12 @K= 20: 22.07
ADE-8/12: 0.21FDE-8/12: 0.39
ADE-8/12: 0.31FDE-8/12: 0.65
ADE-8/12 @K = 20: 7.12FDE-8/12 @K= 20: 11.39