ABSTRACT
Despite the great progress in self-driving technology, it is still challenging to study the safety of self-driving vehicles. Trajectory prediction is one of the core functions of self-driving vehicles. In order to solve the problem of vehicle trajectory prediction, we design a new model framework, which first vectorizes the road structure and agents on the road, and proposes the Transformer encoder to encode the road and the historical trajectory; the interaction module is proposed to extract the interaction features that exist between the target vehicle and the surrounding vehicles and roads; the decoder uses residual connections, and the multi-head attention mechanism extracts the weights canact in the decoder to increase the accuracy of prediction. Experiments are conducted on the HighD public dataset, and the proposed prediction network has a more favorable performance compared to the current mainstream prediction algorithms.
- Huang Y, Du J, Yang Z, A survey on trajectory-prediction methods for autonomous driving [J]. IEEE Transactions on Intelligent Vehicles, 2022, 7(3): 652-674.Google ScholarCross Ref
- Zhang K, Zhao L, Dong C, AI-TP: Attention-based interaction-aware trajectory prediction for autonomous driving [J]. IEEE Transactions on Intelligent Vehicles, 2022, 8(1): 73-83.Google ScholarCross Ref
- Deo N, Rangesh A, Trivedi M M. How would surround vehicles move [J]. A unified framework for maneuver classification and motion prediction, 2018.Google ScholarCross Ref
- Qu R, Huang S, Zhou J, The Vehicle Trajectory Prediction Based on ResNet and EfficientNet Model[J]. 2022.Google Scholar
- Liang M, Yang B, Hu R, Learning lane graph representations for motion forecasting[C]//Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part II 16. Springer International Publishing, 2020, 541-556.Google Scholar
- Gao J, Sun C, Zhao H, Vectornet: Encoding hd maps and agent dynamics from vectorized representation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020, 11525-11533.Google Scholar
- Mercat J, Gilles T, Zoghby N E, Multi-Head Attention for Multi-Modal Joint Vehicle Motion Forecasting [J]. 2019.Google Scholar
- Gehring J, Auli M, Grangier D, Convolutional sequence to sequence learning[C]//International conference on machine learning. PMLR, 2017, 1243-1252.Google Scholar
- Li K, Eiffert S, Shan M, Attentional-GCNN: Adaptive pedestrian trajectory prediction towards generic autonomous vehicle use cases[C]//2021 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2021, 14241-14247.Google Scholar
- Krajewski R, Bock J, Kloeker L, The highd dataset: A drone dataset of naturalistic vehicle trajectories on german highways for validation of highly automated driving systems[C]//2018 21st international conference on intelligent transportation systems (ITSC). IEEE, 2018, 2118-2125.Google Scholar
- Chandra R, Guan T, Panuganti S, Forecasting trajectory and behavior of road-agents using spectral clustering in graph-lstms [J]. IEEE Robotics and Automation Letters, 2020, 5(3): 4882-4890.Google ScholarCross Ref
Index Terms
- Research on the prediction method of vehicle trajectory based on Trasnsformer modeling
Recommendations
Data-Driven Vehicle Trajectory Prediction
SIGSIM-PADS '16: Proceedings of the 2016 ACM SIGSIM Conference on Principles of Advanced Discrete SimulationVehicle trajectory or route prediction is useful in online, data-driven transportation simulation to predict future traffic patterns and congestion, among other uses. The various approaches to route prediction have varying degrees of data required to ...
Vehicle Trajectory Prediction and Cut-In Collision Warning Model in a Connected Vehicle Environment
Side collisions caused by sudden vehicle cut-ins comprise a significant proportion of traffic accidents. Due to the complex and dynamic nature of traffic environments, the warning algorithms in advanced driving assistant systems (ADAS) often misjudge and ...
Prediction of lane change trajectory of autonomous vehicles based on vehicle-environment data fusion
ICBICC '22: Proceedings of the 2022 International Conference on Big Data, IoT, and Cloud ComputingAbstract. Vehicle lane changing is a current research hotspot in the field of autonomous driving. However, the current research does not consider the environmental data of a large number of vehicles as well as the lane change decision risk and lane ...
Comments