List your company

Trajectory Prediction on Safety-critical vehicle-pedestrian interactions

0.054RMSE v_veh^y (m/s)Refined MA-SST-DDPG
0.03240.17820.3240.4698May 17, 2026

Evaluation results

MethodLinks
Refined MA-SST-DDPGModel Type=Multi-Agent Reinforcement Learning, Training Setting=Hybrid Learning2026.050.0540.0240.0530.0310.0960.1140.1020.1120.0720.142
Refined MA-Transformer-DDPGModel Type=Multi-Agent Reinforcement Learning, Training Setting=Hybrid Learning2026.050.0560.0240.0560.0350.1340.1340.1360.1330.0940.194
MA-SST-DDPGModel Type=Multi-Agent Reinforcement Learning, Training Setting=Data-Driven Learning2026.050.0580.0230.0560.030.0980.1180.1030.1260.0780.156
MA-Transformer-DDPGModel Type=Multi-Agent Reinforcement Learning, Training Setting=Data-Driven Learning2026.050.0610.0260.0570.0350.1450.1360.1460.1350.0980.196
Refined MA-DDPGModel Type=Multi-Agent Reinforcement Learning, Training Setting=Hybrid Learning2026.050.1330.0410.1310.0430.3120.1720.2180.1880.1570.305
MA-DDPGModel Type=Multi-Agent Reinforcement Learning, Training Setting=Data-Driven Learning2026.050.1470.0450.1340.0480.3220.1840.2270.2030.1640.328
Refined Transformer-DDPGModel Type=Single-Agent Reinforcement Learning, Training Setting=Hybrid Learning2026.050.2490.0750.2290.0850.3630.3860.4310.3240.2570.505
Transformer-DDPGModel Type=Single-Agent Reinforcement Learning, Training Setting=Data-Driven Learning2026.050.2640.0790.2330.0860.3830.3870.4340.3460.2710.542
Refined DDPGModel Type=Single-Agent Reinforcement Learning, Training Setting=Hybrid Learning2026.050.5570.3060.7310.1820.9140.9610.5990.6220.6231.273
DDPGModel Type=Single-Agent Reinforcement Learning, Training Setting=Data-Driven Learning2026.050.5940.3130.8020.1811.0011.050.6560.6710.6791.357