Motion Estimation on nuScenes Static
0.0041Mean ErrorHPLFlowNet
Evaluation results
| Method | Links | ||
|---|---|---|---|
| HPLFlowNetTraining Protocol=Supervised on nuScenes, Inference Time=0.352s2021.04 | 0.0041 | 0.0002 | |
| MotionNet+MGDATraining Protocol=Supervised on nuScenes, Inference Time=0.019s2021.04 | 0.0201 | 0 | |
| PointRCNNTraining Protocol=Supervised on nuScenes, Inference Time=0.201s2021.04 | 0.0204 | 0 | |
| MotionNetTraining Protocol=Supervised on nuScenes, Inference Time=0.019s2021.04 | 0.0239 | 0 | |
| Self-Supervised Pillar Motion LearningTraining Protocol=Fine-tuned on nuScenes, Inference Time=0.020s2021.04 | 0.0245 | 0 | |
| MotionNet (pillar-based)Training Protocol=Supervised on nuScenes, Inference Time=0.019s2021.04 | 0.0258 | — | |
| LSTMEncoderDecoderTraining Protocol=Supervised on nuScenes, Inference Time=0.042s2021.04 | 0.0358 | 0 | |
| FlowNet3DTraining Protocol=Supervised on nuScenes, Inference Time=0.434s2021.04 | 0.041 | 0 | |
| Self-Supervised Pillar Motion LearningTraining Protocol=Self-supervised, Inference Time=0.020s2021.04 | 0.162 | 0.001 | |
| FlowNet3DTraining Protocol=Pre-trained on FlyingThings3D and KITTI, Inference Time=0.434s2021.04 | 2.0514 | 0 | |
| HPLFlowNetTraining Protocol=Pre-trained on FlyingThings3D and KITTI, Inference Time=0.352s2021.04 | 2.2165 | 1.4925 |