如何实现典型交通参与者行为预测？

CoverNet: 多模态行为预测使用轨迹集Diverse and Admissible Trajectory Forecasting through Multimodal Context UnderstandingDenseTNT: 端到端轨迹预测从密集目标集TNT: 目标驱动N轨迹预测VectorNet: 从向量化表示编码高清地图和代理动态LOKI: 长期和关键意图轨迹预测FIERY: 从周围单目摄像头预测未来实例鸟瞰图Lane-Attention: 通过学习车道上的注意力预测车辆移动轨迹LaneRCNN: 图为中心运动预测的分布式表示（CVPR2021）LaPred: 车道感知多模态未来动态代理轨迹预测Learning Lane Graph Representations for Motion Forecasting (ECCV2020)Learning to Predict Vehicle Trajectories with Model-based PlanningOpen-set Intersection Intention Prediction for Autonomous DrivingPiP: 自动驾驶规划信息轨迹预测Probabilistic Multi-modal Trajectory Prediction with Lane Attention for Autonomous VehiclesRAIN: 强化混合注意力推理网络运动预测Spatial-Channel Transformer Network for Trajectory Prediction on the Traffic ScenesTPCN: 运动预测的时间点云网络TPNet: 轨迹提议网络运动预测Unlimited Neighborhood Interaction for Heterogeneous Trajectory PredictionHeterogeneous Edge-Enhanced Graph Attention Network For Multi-Agent Trajectory PredictionImplicit Latent Variable Model for Scene-Consistent Motion ForecastingConvolutional Social Pooling for Vehicle Trajectory PredictionAttention Based Vehicle Trajectory PredictionVehicle Trajectory Prediction Using LSTMs with Spatial–Temporal Attention MechanismsA Dynamic and Static Context-Aware Attention Network for Trajectory PredictionSocial LSTM: 拥挤空间人类轨迹预测SocialGAN: 社交可接受轨迹生成对抗网络Hierarchical Adaptable and Transferable Networks (HATN) for Driving Behavior PredictionMulti-Agent Driving Behavior Prediction across Different Scenarios with Self-Supervised Domain KnowledgeAgentFormer: 社会时间多代理预测的代理感知变换器M2I: 从分解边际轨迹预测到交互预测MUSE-VAE: 环境感知长期轨迹预测的多尺度VAEIdentifying Driver Interactions via Conditional Behavior PredictionRemember Intentions: 基于回顾记忆的轨迹预测GRIP++: 增强基于图的交互感知自动驾驶轨迹预测MANTRA: 多轨迹预测的记忆增强网络在自动驾驶和智能交通系统领域，预测交通参与者的行为至关重要。通过分析轨迹集、多模态上下文理解、目标驱动预测、向量化表示、长期意图识别、车道注意力机制、图中心运动预测、模型基础规划、开放集交叉意图预测、规划信息轨迹预测、概率多模态轨迹预测、强化混合注意力推理网络、空间通道变换器网络、时间点云网络、轨迹提议网络、无限邻域交互、异构边缘增强图注意力网络、隐式潜在变量模型、卷积社会池化、基于注意力的车辆轨迹预测、基于LSTM的空间-时间注意力机制、动态静态上下文感知注意力网络、社交LSTM、社交GAN、层次适应和可转移网络、多代理驾驶行为预测、代理感知变换器、从分解边际轨迹预测到交互预测、环境感知长期轨迹预测的多尺度VAE、通过条件行为预测识别驾驶员交互、基于回顾记忆的轨迹预测、增强基于图的交互感知自动驾驶轨迹预测、多轨迹预测的记忆增强网络等方法，可以有效地预测交通参与者的行为，提高自动驾驶的安全性和效率。

CoverNet:MultimodalBehaviorPredictionusingTrajectorySets

DiverseandAdmissibleTrajectoryForecastingthroughMultimodalContextUnderstanding

DenseTNT:End-to-endTrajectoryPredictionfromDenseGoalSets

TNT:Target-driveNTrajectoryPrediction

VectorNet:EncodingHDMapsandAgentDynamicsfromVectorizedRepresentation

LOKI:LongTermandKeyIntentionsforTrajectoryPrediction

FIERY:FutureInstancePredictioninBird’s-EyeViewfromSurroundMonocularCameras

Lane-Attention:PredictingVehicles’MovingTrajectoriesbyLearningTheirAttentionOverLanes

LaneRCNN:DistributedRepresentationsforGraph-CentricMotionForecasting(CVPR2021)

LaPred:Lane-AwarePredictionofMulti-ModalFutureTrajectoriesofDynamicAgents

LearningLaneGraphRepresentationsforMotionForecasting(ECCV2020)

LearningtoPredictVehicleTrajectorieswithModel-basedPlanning

Open-setIntersectionIntentionPredictionforAutonomousDriving

PiP:Planning-informedTrajectoryPredictionforAutonomousDriving

ProbabilisticMulti-modalTrajectoryPredictionwithLaneAttentionforAutonomousVehicles

RAIN:ReinforcedHybridAttentionInferenceNetworkforMotionForecasting

Spatial-ChannelTransformerNetworkforTrajectoryPredictionontheTrafficScenes

TPCN:TemporalPointCloudNetworksforMotionForecasting

TPNet:TrajectoryProposalNetworkforMotionPrediction

UnlimitedNeighborhoodInteractionforHeterogeneousTrajectoryPrediction

HeterogeneousEdge-EnhancedGraphAttentionNetworkForMulti-AgentTrajectoryPrediction

ImplicitLatentVariableModelforScene-ConsistentMotionForecasting

ConvolutionalSocialPoolingforVehicleTrajectoryPrediction

AttentionBasedVehicleTrajectoryPrediction

VehicleTrajectoryPredictionUsingLSTMsWithSpatial–TemporalAttentionMechanisms

ADynamicandStaticContext-AwareAttentionNetworkforTrajectoryPrediction

SocialLSTM:HumanTrajectoryPredictioninCrowdedSpaces

SocialGAN:SociallyAcceptableTrajectorieswithGenerativeAdversarialNetworks

HierarchicalAdaptableandTransferableNetworks(HATN)forDrivingBehaviorPrediction

Multi-AgentDrivingBehaviorPredictionacrossDifferentScenarioswithSelf-SupervisedDomainKnowledge

AgentFormer:Agent-AwareTransformersforSocio-TemporalMulti-AgentForecasting

M2I:FromFactoredMarginalTrajectoryPredictiontoInteractivePrediction

MUSE-VAE:Multi-ScaleVAEforEnvironment-AwareLongTermTrajectoryPrediction

IdentifyingDriverInteractionsviaConditionalBehaviorPrediction

RememberIntentions:Retrospective-Memory-basedTrajectoryPrediction

GRIP++:EnhancedGraph-basedInteraction-awareTrajectoryPredictionforAutonomousDriving

MANTRA:MemoryAugmentedNetworksforMultipleTrajectoryPrediction

以上内容由58汽车提供。如有任何买车、用车、养车、玩车相关问题，欢迎在下方表单填写您的信息，我们将第一时间与您联系，为您提供快捷、实用、全面的解决方案。