Convolutional Social Pooling for Vehicle Trajectory Prediction

By Nachiket Deo, Mohan Manubhai Trivedi. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2018

1,014 citations recorded by the source corpus.

Forecasting the motion of surrounding vehicles is a critical ability for an autonomous vehicle deployed in complex traffic. Motion of all vehicles in a scene is governed by the traffic context, i.e., the motion and relative spatial configuration of neighboring vehicles. In this paper we propose an LSTM encoder-decoder model that uses convolutional social pooling as an improvement to social pooling layers for robustly learning interdependencies in vehicle motion. Additionally, our model outputs a multi-modal predictive distribution over future trajectories based on maneuver classes. We evaluate our model using the publicly available NGSIM US-101 and I-80 datasets. Our results show improvement over the state of the art in terms of RMS values of prediction error and negative log-likelihoods of true future trajectories under the model's predictive distribution. We also present a qualitative analysis of the model's predicted distributions for various traffic scenarios.

Subfields: Automotive Engineering, Building and Construction, Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/7f9832031d61c46f051dad9a0d421338888b6b00

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Automotive Engineering

In Building and Construction

In Control and Systems Engineering

Citations in the corpus

PIE: A Large-Scale Dataset and Models for Pedestrian Intention Estimation and Trajectory Prediction (2019, background): “Action (or behavior) prediction algorithms may take different forms such as generating future frames [20, 19, 24, 6], predicting the type of action [15, 21, 7], measuring confidence in the occurrence of an event [27, 37, 10], and forecasting the motion of objects [25, 40, 43, 1, 17, 5, 8].”
A Survey on Trajectory-Prediction Methods for Autonomous Driving (2022, background): “Unlike the social pooling proposed in [102], the pooling module used here deals with the interaction between pedestrians.”
Deep Learning-Based Vehicle Behavior Prediction for Autonomous Driving Applications: A Review (2020, background): “[38] The representation in [37]”
Query-Centric Trajectory Prediction (2023, methodology): “Anchor-based methods achieve this with the guidance of anchors, which facilitate multimodal prediction by leveraging predefined maneuvers [12], candidate trajectories [6, 39], or map-adaptive goals [53, 55].”
GRIP: Graph-based Interaction-aware Trajectory Prediction (2019, methodology): “In this subsection, we compare our proposed scheme to the following baselines (as done in [11]) and some existing solutions:”
CoverNet: Multimodal Behavior Prediction Using Trajectory Sets (2019, background): “Multimodal approaches output either a distribution over multiple trajectories [7, 13, 20, 14] or a spatial-temporal occupancy map [20, 28, 34].”
Holistic Network Virtualization and Pervasive Network Intelligence for 6G (2023, methodology): “for end users by processing AI tasks locally, such as next word prediction in mobile keyboards [148], user traffic demand prediction [149], and vehicle trajectory prediction [150].”
HiVT: Hierarchical Vector Transformer for Multi-Agent Motion Prediction (2022, background): “Since social interactions are ubiquitous in traffic scenarios and significantly impact the future motions of traffic agents, many motion prediction approaches have considered the dependencies between agents’ behaviors and reasoned agent-agent interactions using social pooling [2,13,23], graph neural”
Multimodal Motion Prediction with Stacked Transformers (2021, methodology): “Recently deep learning-based methods have been widely used for motion prediction [1, 14, 8, 2, 11].”
Lidar for Autonomous Driving: The Principles, Challenges, and Trends for Automotive Lidar and Perception Systems (2020, background): “The authors in [43] and [44] applied and improved the original social-pooling part for the purpose of vehicle-trajectory prediction.”
Conditional Vehicle Trajectories Prediction in CARLA Urban Environment (2019, methodology): “The dimensions of the area around the ego, used in [9], corresponds to 3 standard lanes side by side for around 2s front and rear at 50kph.”
Attention Based Vehicle Trajectory Prediction (2020, methodology): “[7] extend the social pooling and deploy it for vehicle trajectory prediction task.”
Argoverse: 3D Tracking and Forecasting With Rich Maps (2019, background): “[8] propose a convolutional social pooling approach wherein”
Multi-Agent Tensor Fusion for Contextual Trajectory Prediction (2019, methodology): “As in [9], we report the Root Mean Square Error in meters with respect to each timestep t within the prediction horizon: RMSE ( t ) = , where n is the total number of agents in the validation set, x it denotes the x coordinate of the i -th car in the dataset at future timestep t , and y it the y…”
TraPHic: Trajectory Prediction in Dense and Heterogeneous Traffic Using Weighted Interactions (2018, methodology): “• Convolutional-Social-LSTM (CS-LSTM): A variant of SLSTM adding convolutions to the network in [3] in order to predict trajectories in sparse highway traffic [6].”