REVIEW 16 cited by
Scene Transformer: A unified architecture for predicting multiple agent trajectories
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Predicting the motion of multiple agents is necessary for planning in dynamic environments. This task is challenging for autonomous driving since agents (e.g. vehicles and pedestrians) and their associated behaviors may be diverse and influence one another. Most prior work have focused on predicting independent futures for each agent based on all past motion, and planning against these independent predictions. However, planning against independent predictions can make it challenging to represent the future interaction possibilities between different agents, leading to sub-optimal planning. In this work, we formulate a model for predicting the behavior of all agents jointly, producing consistent futures that account for interactions between agents. Inspired by recent language modeling approaches, we use a masking strategy as the query to our model, enabling one to invoke a single model to predict agent behavior in many ways, such as potentially conditioned on the goal or full future trajectory of the autonomous vehicle or the behavior of other agents in the environment. Our model architecture employs attention to combine features across road elements, agent interactions, and time steps. We evaluate our approach on autonomous driving datasets for both marginal and joint motion prediction, and achieve state of the art performance across two popular datasets. Through combining a scene-centric approach, agent permutation equivariant model, and a sequence masking strategy, we show that our model can unify a variety of motion prediction tasks from joint motion predictions to conditioned prediction.
Forward citations
Cited by 16 Pith papers
-
Scaling Laws of Motion Forecasting and Planning -- Technical Report
Motion forecasting models improve with compute as a power law, with optimal model size growing 1.5x faster than dataset size, and closed-loop driving failures also decreasing with scale.
-
AutoWorld: Learning Multi-Agent Traffic Simulation with Self-Supervised World Models
AutoWorld learns a self-supervised LiDAR occupancy world model and conditions a diffusion-based motion generator on its forecasts, reporting the top Waymo Sim Agents realism score.
-
To New Beginnings: A Survey of Unified Perception in Autonomous Vehicle Software
A taxonomy that classifies unified perception methods in autonomous driving into Early, Late, and Full Unified Perception based on task integration, tracking formulation, and representation flow.
-
Goal-based Trajectory Prediction for improved Cross-Dataset Generalization
A graph neural network with staged lane-then-point goal selection cuts the cross-dataset error drop (e.g., b-minFDE6 relative drop 8% vs 20-30%) when moving from Argoverse2 to NuScenes.
-
ILNet: Trajectory Prediction with Inverse Learning Attention for Enhancing Intention Capture
ILNet reports top INTERACTION joint metrics and strong Argoverse marginal metrics using inverse temporal attention plus learned dynamic anchor refinement.
-
LTMSformer: A Local Trend-Aware Attention and Motion State Encoding Transformer for Multi-Agent Trajectory Prediction
LTMSformer improves multi-agent trajectory prediction on Argoverse 1 by adding local trend-aware attention, high-order motion state encoding, and an MLP-based refinement module to the HiVT architecture.
-
TrajFlow: Multi-modal Motion Prediction via Flow Matching
TrajFlow uses flow matching with a multi-query transformer to predict multiple trajectories in one pass and a Plackett-Luce ranking loss to improve confidence scores, reporting small SOTA gains on WOMD.
-
A Dynamic Scene Interaction Reasoning Framework for Scene-level Lane-Change Intention and Trajectory Prediction of Multiple Interacting Vehicles
A dynamic graph-attention model jointly predicts every nearby vehicle’s lane-change intention and trajectory, cutting trajectory error by up to ~53% and improving scene coherence on NGSIM and highD.
-
Pivot-Centric Trajectory Prediction: Bridging Long Horizons via Dynamical Guidance
PCTP predicts multiple hierarchical "pivot" waypoints first and then refines the segments between them, improving long-horizon trajectory prediction on Argoverse I and II.
-
TGRIP: A Text-Guided Approach to Vehicle Instance Prediction in Autonomous Driving
Auxiliary CLIP-derived BEV semantic supervision during training improves nuScenes end-to-end vehicle instance prediction over a geometric-only baseline, with the semantic head removed at inference.
-
From Camera-Based Sensing to Reasoning: A Comprehensive Review Toward Proactive Vulnerable Road User Safety
A survey organizing recent camera-based AI methods for vulnerable road user safety into four interlocking visual tasks and four open deployment challenges.
-
Adaptive Output Steps: FlexiSteps Network for Dynamic Trajectory Prediction
FSN dynamically selects the number of future trajectory steps to predict, using a learned classifier plus a Fréchet-distance-based score, claiming improved accuracy and efficiency on two driving benchmarks.
-
Foresight in Motion: Reinforcing Trajectory Prediction with Reward Heuristics
FiM predicts future trajectories by first learning a reward distribution over a grid world via inverse reinforcement learning, then rolling out intention plans that condition a Mamba-enhanced trajectory decoder.
-
RoCA: Robust Cross-Domain End-to-End Autonomous Driving
RoCA, a Gaussian-process codebook over ego and agent tokens, improves cross-domain generalization and adaptation of end-to-end autonomous driving models without extra inference cost.
-
Goal-conditioned Hierarchical Reinforcement Learning for Sample-efficient and Safe Autonomous Driving at Intersections
A hierarchical RL agent with a goal-conditioned collision prediction module achieves 94.7% success and 3.3% collisions in SMARTS intersection tasks, outperforming flat RL baselines.
-
PhysVarMix: Physics-Informed Variational Mixture Model for Multi-Modal Trajectory Prediction
PhysVarMix is a claimed physics-informed variational mixture model for multimodal trajectory prediction, but its latent variable never affects the predicted trajectories, so it reduces to a standard mixture density ne...
Discussion (0). Sign in to comment.