mlsanityData to 14 Oct 2026

Bootstrap Motion Forecasting With Self-Consistent Constraints

arXiv 2204.05859 ↗First posted 12 Apr 2022cs.CV cs.ROcited by 23

Not on the map. Why?

Abstract (arXiv)

We present a novel framework to bootstrap Motion forecasting with Self-consistent Constraints (MISC). The motion forecasting task aims at predicting future trajectories of vehicles by incorporating spatial and temporal information from the past. A key design of MISC is the proposed Dual Consistency Constraints that regularize the predicted trajectories under spatial and temporal perturbation during training. Also, to model the multi-modality in motion forecasting, we design a novel self-ensembling scheme to obtain accurate teacher targets to enforce the self-constraints with multi-modality supervision. With explicit constraints from multiple teacher targets, we observe a clear improvement in the prediction performance. Extensive experiments on the Argoverse motion forecasting benchmark and Waymo Open Motion dataset show that MISC significantly outperforms the state-of-the-art methods. As the proposed strategies are general and can be easily incorporated into other motion forecasting approaches, we also demonstrate that our proposed scheme consistently improves the prediction performance of several existing methods.

Key facts. Cited by 23 papers in the map (data through 14 Oct 2026). Builds on 40 of its 66 references that are in the map. First posted on arXiv 12 Apr 2022.

Builds on

40 of 66 references.

References · 40

  1. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  2. 2019Scalability in Perception for Autonomous Driving: Waymo Open Datasetcited by 2,451
  3. 2017Cascade R-CNN: Delving into High Quality Object Detectioncited by 1,276
  4. 2016Temporal Ensembling for Semi-Supervised Learningcited by 1,057
  5. 2017Unsupervised Learning of Depth and Ego-Motion from Videocited by 782
  6. 2015Convolutional Networks on Graphs for Learning Molecular Fingerprintscited by 754
  7. 2019Argoverse: 3D Tracking and Forecasting with Rich Mapscited by 732
  8. 2021Large Scale Interactive Motion Forecasting for Autonomous Driving : The Waymo Open Motion Datasetcited by 537
  9. 2020VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representationcited by 530
  10. 2016Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learningcited by 441
  11. 2019MultiPath: Multiple Probabilistic Anchor Trajectory Hypotheses for Behavior Predictioncited by 420
  12. 2020Learning Lane Graph Representations for Motion Forecastingcited by 382
28 more
  1. 2020TNT: Target-driveN Trajectory Predictioncited by 347
  2. 2021DenseTNT: End-to-end Trajectory Prediction from Dense Goal Setscited by 295
  3. 2019Learning Correspondence from the Cycle-Consistency of Timecited by 269
  4. 2019CoverNet: Multimodal Behavior Prediction using Trajectory Setscited by 259
  5. 2015Deep Convolutional Networks on Graph-Structured Datacited by 246
  6. 2021MultiPath++: Efficient Information Fusion and Trajectory Aggregation for Behavior Predictioncited by 246
  7. 2021End-to-end Interpretable Neural Motion Plannercited by 242
  8. 2014Learning with Pseudo-Ensemblescited by 241
  9. 2020It Is Not the Journey but the Destination: Endpoint Conditioned Trajectory Predictioncited by 236
  10. 2021Multimodal Motion Prediction with Stacked Transformerscited by 208
  11. 2021IntentNet: Learning to Predict Intention from Raw Sensor Datacited by 158
  12. 2021GOHOME: Graph-Oriented Heatmap Output for future Motion Estimationcited by 146
  13. 2021HOME: Heatmap Output for future Motion Estimationcited by 122
  14. 2021SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloudcited by 119
  15. 2021LaneRCNN: Distributed Representations for Graph-Centric Motion Forecastingcited by 113
  16. 2018Baidu Apollo EM Motion Plannercited by 100
  17. 2016Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypothesescited by 97
  18. 2019Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Predictioncited by 96
  19. 2021TPCN: Temporal Point Cloud Networks for Motion Forecastingcited by 94
  20. 2019Multi-Head Attention for Multi-Modal Joint Vehicle Motion Forecastingcited by 87
  21. 2016Stochastic Multiple Choice Learning for Training Diverse Deep Ensemblescited by 77
  22. 2020Blind Video Temporal Consistency via Deep Video Priorcited by 76
  23. 2021Learning to Predict Vehicle Trajectories with Model-based Planningcited by 63
  24. 2021Divide-and-Conquer for Lane-Aware Diverse Trajectory Predictioncited by 37
  25. 2021Dual-Camera Super-Resolution with Aligned Attention Modulescited by 34
  26. 2021Internal Video Inpainting by Implicit Long-range Propagationcited by 19
  27. 2018Interaction-Aware Probabilistic Behavior Prediction in Urban Environmentscited by 7
  28. 2019A Hierarchical Network for Diverse Trajectory Proposalscited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 3; 2023: 8; 2024: 7; 2025: 3; 2026: 2.

23 papers.

Sort:

By area:

  1. 2022GANet: Goal Area Network for Motion Forecastingcited by 49
  2. 2023ProphNet: Efficient Agent-Centric Motion Forecasting with Anchor-Informed Proposalscited by 48
  3. 2024SmartRefine: A Scenario-Adaptive Refinement Framework for Efficient Motion Predictioncited by 34
  4. 2022SSL-Lanes: Self-Supervised Learning for Motion Forecasting in Autonomous Drivingcited by 32
  5. 2024Multi-agent Long-term 3D Human Pose Forecasting via Interaction-aware Trajectory Conditioningcited by 30
  6. 2023Dynamic Scenario Representation Learning for Motion Forecasting with Heterogeneous Graph Convolutional Recurrent Networkscited by 28
  7. 2023SEPT: Towards Efficient Scene Representation Learning for Motion Predictioncited by 27
  8. 2023What Truly Matters in Trajectory Prediction for Autonomous Driving?cited by 17
  9. 2023Bridging the Gap Between Multi-Step and One-Shot Trajectory Prediction via Self-Supervisioncited by 7
  10. 2024SmartPretrain: Model-Agnostic and Dataset-Agnostic Representation Learning for Motion Predictioncited by 7
  11. 2024FutureNet-LOF: Joint Trajectory Prediction and Lane Occupancy Field Prediction with Future Context Encodingcited by 6
  12. 2024Learning Online Belief Prediction for Efficient POMDP Planning in Autonomous Drivingcited by 5
11 more
  1. 2024Self-Supervised Class-Agnostic Motion Prediction with Spatial and Temporal Consistency Regularizationscited by 5
  2. 2023TSGN: Temporal Scene Graph Neural Networks with Projected Vectorized Representation for Multi-Agent Motion Predictioncited by 2
  3. 2024Towards Practical Human Motion Prediction with LiDAR Point Cloudscited by 2
  4. 2026Recover to Predict: Progressive Retrospective Learning for Variable-Length Trajectory Predictioncited by 2
  5. 2022Improving Motion Forecasting for Autonomous Driving with the Cycle Consistency Losscited by 1
  6. 2023Improving Trajectory Prediction in Dynamic Multi-Agent Environment by Dropping Waypointscited by 1
  7. 2023Enhancing Trajectory Prediction through Self-Supervised Waypoint Noise Predictioncited by 1
  8. 2025ILNet: Trajectory Prediction with Inverse Learning Attention for Enhancing Intention Capturecited by 0
  9. 2025Ensemble of Pre-Trained Models for Long-Tailed Trajectory Predictioncited by 0
  10. 2025V2X-RECT: An Efficient V2X Trajectory Prediction Framework via Redundant Interaction Filtering and Tracking Error Correctioncited by 0
  11. 2026Pivot-Centric Trajectory Prediction: Bridging Long Horizons via Dynamical Guidancecited by 0