Trajectory Prediction on Stanford Drone Dataset (SDD)
7.37ADETIGFlow-GRPO
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| TIGFlow-GRPOMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 7.37 | 11.67 | — | — | |
| MIDEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 7.6 | 14.3 | — | — | |
| MoFlowMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 7.63 | 12.25 | — | — | |
| LMTraj-SUPEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 7.8 | 10.1 | — | — | |
| NPSNEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 7.9 | 11.9 | — | — | |
| EigenTrajMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 8.05 | 13.25 | — | — | |
| EigenTrajectoryEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 8.1 | 13.1 | — | — | |
| EqMotionEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 8.1 | 11.7 | — | — | |
| LEDMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 8.49 | 12.67 | — | — | |
| LEDEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 8.5 | 11.7 | — | — | |
| SocialVAEEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 8.6 | 11.9 | — | — | |
| AgentFormerEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 8.7 | 14.9 | — | — | |
| GP-GraphEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 9.1 | 13.8 | — | — | |
| GroupNetMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 9.31 | 16.11 | — | — | |
| MIDMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 9.73 | 15.32 | — | — | |
| PECNetEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 10 | 15.9 | — | — | |
| Trajectron+++Evaluation Mode=Stochastic, Number of samples (K)=202024.03 | 11.4 | 20.1 | — | — | |
| Y-netMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 11.51 | 20.24 | — | — | |
| Social-GANEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 13.6 | 24.6 | — | — | |
| LMTraj-SUPEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 17.5 | 34.5 | — | — | |
| EigenTrajectoryEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 20.7 | 41.9 | — | — | |
| Social-STGCNNEvaluation Mode=Stochastic, Number of samples (K)=202024.03 | 20.8 | 33.2 | — | — | |
| NPSNEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 22.1 | 38 | — | — | |
| Trajectron+++Evaluation Mode=Deterministic, Number of samples (K)=12024.03 | 22.7 | 42 | — | — | |
| SocialVAEEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 24.2 | 49.3 | — | — | |
| GP-GraphEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 24.7 | 49 | — | — | |
| MIDEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 25.2 | 57.6 | — | — | |
| SocialGANMetric Type=best-of-K, Evaluation Setting=image-plane2026.03 | 27.23 | 41.44 | — | — | |
| Social-GANEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 27.3 | 41.4 | — | — | |
| STGATEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 28 | 41.3 | — | — | |
| STAR-DEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 28.8 | 51.4 | — | — | |
| PECNetEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 29.8 | 65.1 | — | — | |
| Social-LSTMEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 31.2 | 57 | — | — | |
| SR-LSTMEvaluation Mode=Deterministic, Number of samples (K)=12024.03 | 31.4 | 56.8 | — | — | |
| AGMAVenue=(Ours)2026.02 | — | — | 7.23 | 10.92 | |
| EigenTrajVenue=ICCV’232026.02 | — | — | 8.05 | 13.25 | |
| MGFVenue=NeurIPS’242026.02 | — | — | 7.74 | 12.07 | |
| MoFlowVenue=CVPR’252026.02 | — | — | 7.66 | 12.39 | |
| NMRFVenue=ICLR’252026.02 | — | — | 7.2 | 11.29 | |
| TUTRVenue=ICCV’232026.02 | — | — | 7.76 | 12.69 |