Bootstrap Motion Forecasting With Self-Consistent Constraints
Authors:
- 1 October 2023
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 8470-8480
Abstract
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.Keywords
All Related Versions
This publication has 41 references indexed in Scilit:
- It Is Not the Journey But the Destination: Endpoint Conditioned Trajectory Prediction2020
- Learning Lane Graph Representations for Motion Forecasting2020
- Scalability in Perception for Autonomous Driving: Waymo Open Dataset2020
- Argoverse: 3D Tracking and Forecasting With Rich Maps2019
- Learning Correspondence From the Cycle-Consistency of Time2019
- End-To-End Interpretable Neural Motion Planner2019
- Overcoming Limitations of Mixture Density Networks: A Sampling and Fitting Framework for Multimodal Future Prediction2019
- Cascade R-CNN: Delving Into High Quality Object Detection2018
- Unsupervised Learning of Depth and Ego-Motion from Video2017
- Ensemble Methods in Machine Learning2000