FJMP: Factorized Joint Multi-Agent Motion Prediction over Learned Directed Acyclic Interaction Graphs
Origin paper
FJMP: Factorized Joint Multi-Agent Motion Prediction over Learned Directed Acyclic Interaction Graphs
MacFormer: Map-Agent Coupled Transformer for Real-Time and Robust Trajectory Prediction
Forecast-MAE: Self-supervised Pre-training for Motion Forecasting with Masked Autoencoders
ADAPT: Efficient Multi-Agent Trajectory Prediction with Adaptation
R-Pred: Two-Stage Motion Prediction Via Tube-Query Attention-Based Trajectory Refinement
Leveraging Future Relationship Reasoning for Vehicle Trajectory Prediction
Towards Capturing the Temporal Dynamics for Trajectory Prediction: a Coarse-to-Fine Approach
BiFF: Bi-level Future Fusion with Polyline-based Coordinate for Interactive Trajectory Prediction
MTR++: Multi-Agent Motion Prediction With Symmetric Scene Modeling and Guided Intention Querying
RMP: A Random Mask Pretrain Framework for Motion Prediction
Two-Stage Context-Aware model for Predicting Future Motion of Dynamic Agents
Query-Centric Trajectory Prediction
Tracing the Influence of Predecessors on Trajectory Prediction
HDGT: Heterogeneous Driving Graph Transformer for Multi-Agent Trajectory Prediction via Scene Encoding
PBP: Path-based Trajectory Prediction for Autonomous Driving
Exploiting map information for self-supervised learning in motion forecasting
HPNet: Dynamic Trajectory Forecasting with Historical Prediction Attention
Real-Time Motion Prediction via Heterogeneous Polyline Transformer with Relative Pose Encoding
GANet: Goal Area Network for Motion Forecasting
LAformer: Trajectory Prediction for Autonomous Driving with Lane-Aware Scene Constraints
SemanticFormer: Holistic and Semantic Traffic Scene Representation for Trajectory Prediction using Knowledge Graphs
Traj-MAE: Masked Autoencoders for Trajectory Prediction
ProIn: Learning to Predict Trajectory Based on Progressive Interactions for Autonomous Driving
Towards Trustworthy Multi-Modal Motion Prediction: Evaluation and Interpretability
Motion Transformer with Global Intention Localization and Local Movement Refinement
THOMAS: Trajectory Heatmap Output with learned Multi-Agent Sampling
Learn TAROT with MENTOR: A Meta-Learned Self-supervised Approach for Trajectory Prediction
EDA: Evolving and Distinct Anchors for Multimodal Motion Prediction
ProphNet: Efficient Agent-Centric Motion Forecasting with Anchor-Informed Proposals
M2I: From Factored Marginal Trajectory Prediction to Interactive Prediction
FIMP: Future Interaction Modeling for Multi-Agent Motion Prediction
GoRela: Go Relative for Viewpoint-Invariant Motion Forecasting
Dynamic Scenario Representation Learning for Motion Forecasting With Heterogeneous Graph Convolutional Recurrent Networks
TSGN: Temporal Scene Graph Neural Networks with Projected Vectorized Representation for Multi-Agent Motion Prediction
A Hierarchical Hybrid Learning Framework for Multi-agent Trajectory Prediction
FGNet: A Graph-Based Motion Forecasting Method From a Future Perspective
Heterogeneous Graph-based Trajectory Prediction using Local Map Context and Social Interactions
Trajectory Forecasting on Temporal Graphs
ParallelNet: Multi-mode Trajectory Prediction by Multi-mode Trajectory Fusion
GOHOME: Graph-Oriented Heatmap Output for future Motion Estimation
NEAR: Noise-Aware Temporal Encoder and Adaptive Recurrent Interaction for Motion Forecasting
Log in to saveSave
Predicting the future motion of road agents is a critical task in an autonomous driving pipeline. In this work, we address the problem of generating a set of scene-level, or joint, future trajectory predictions in multi-agent driving scenarios. To this end, we propose FJMP, a Factorized Joint Motion Prediction framework for multi-agent interactive driving scenarios. FJMP models the future scene interaction dynamics as a sparse directed interaction graph, where edges denote explicit interactions between agents. We then prune the graph into a directed acyclic graph (DAG) and decompose the joint prediction task into a sequence of marginal and conditional predictions according to the partial ordering of the DAG, where joint future trajectories are decoded using a directed acyclic graph neural network (DAGNN). We conduct experiments on the INTERACTION and Argoverse 2 datasets and demonstrate that FJMP produces more accurate and scene-consistent joint trajectory predictions than non-factorized approaches, especially on the most interactive and kinematically interesting agents. FJMP ranks 1st on the multi-agent test leaderboard of the INTERACTION dataset.