Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CAMNet: Leveraging Cooperative Awareness Messages for Vehicle Trajectory Prediction

Mattia Grasselli, Angelo Porrello, Carlo Augusto Grazia

TL;DR

The paper tackles trajectory forecasting under limited sensor visibility by leveraging Cooperative Awareness Messages (CAMs) through CAMNet, a graph-based variational recurrent model. CAMNet models inter-vehicle interactions via Graph Neural Networks and learns a multi-modal distribution of future trajectories by optimizing a $eta$-weighted ELBO, enabling multiple plausible paths. The authors introduce two datasets (Argoverse 2 and a CAM-based Modena dataset) and show that CAM data can improve prediction accuracy in multi-agent settings, though performance remains constrained by CAM coverage and the absence of HD-map context. The work highlights the potential of cooperative sensing to extend situational awareness and points to future directions including context integration and vision-based priors to compensate for map absence.

Abstract

Autonomous driving remains a challenging task, particularly due to safety concerns. Modern vehicles are typically equipped with expensive sensors such as LiDAR, cameras, and radars to reduce the risk of accidents. However, these sensors face inherent limitations: their field of view and line of sight can be obstructed by other vehicles, thereby reducing situational awareness. In this context, vehicle-to-vehicle communication plays a crucial role, as it enables cars to share information and remain aware of each other even when sensors are occluded. One way to achieve this is through the use of Cooperative Awareness Messages (CAMs). In this paper, we investigate the use of CAM data for vehicle trajectory prediction. Specifically, we design and train a neural network, Cooperative Awareness Message-based Graph Neural Network (CAMNet), on a widely used motion forecasting dataset. We then evaluate the model on a second dataset that we created from scratch using Cooperative Awareness Messages, in order to assess whether this type of data can be effectively exploited. Our approach demonstrates promising results, showing that CAMs can indeed support vehicle trajectory prediction. At the same time, we discuss several limitations of the approach, which highlight opportunities for future research.

CAMNet: Leveraging Cooperative Awareness Messages for Vehicle Trajectory Prediction

TL;DR

The paper tackles trajectory forecasting under limited sensor visibility by leveraging Cooperative Awareness Messages (CAMs) through CAMNet, a graph-based variational recurrent model. CAMNet models inter-vehicle interactions via Graph Neural Networks and learns a multi-modal distribution of future trajectories by optimizing a -weighted ELBO, enabling multiple plausible paths. The authors introduce two datasets (Argoverse 2 and a CAM-based Modena dataset) and show that CAM data can improve prediction accuracy in multi-agent settings, though performance remains constrained by CAM coverage and the absence of HD-map context. The work highlights the potential of cooperative sensing to extend situational awareness and points to future directions including context integration and vision-based priors to compensate for map absence.

Abstract

Autonomous driving remains a challenging task, particularly due to safety concerns. Modern vehicles are typically equipped with expensive sensors such as LiDAR, cameras, and radars to reduce the risk of accidents. However, these sensors face inherent limitations: their field of view and line of sight can be obstructed by other vehicles, thereby reducing situational awareness. In this context, vehicle-to-vehicle communication plays a crucial role, as it enables cars to share information and remain aware of each other even when sensors are occluded. One way to achieve this is through the use of Cooperative Awareness Messages (CAMs). In this paper, we investigate the use of CAM data for vehicle trajectory prediction. Specifically, we design and train a neural network, Cooperative Awareness Message-based Graph Neural Network (CAMNet), on a widely used motion forecasting dataset. We then evaluate the model on a second dataset that we created from scratch using Cooperative Awareness Messages, in order to assess whether this type of data can be effectively exploited. Our approach demonstrates promising results, showing that CAMs can indeed support vehicle trajectory prediction. At the same time, we discuss several limitations of the approach, which highlight opportunities for future research.
Paper Structure (14 sections, 2 equations, 4 figures, 3 tables)

This paper contains 14 sections, 2 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Problem formulation
  4. Datasets
  5. Argoverse 2 Motion Forecasting Dataset
  6. CAM-based Dataset
  7. Metrics
  8. Proposed Model
  9. Architecture Overview
  10. Encoder, decoder, and prior blocks
  11. Experiments
  12. Argoverse 2 Motion Forecasting Dataset
  13. CAM-based Dataset
  14. Conclusions and Future Works

Figures (4)

  • Figure 1: Illustration of the first interpolation performed. Original CAM data (black) and interpolated values (red) are shown.
  • Figure 2: Brief description of the overall structure and main components of CAMNet.
  • Figure 3: Analysis on the variation of the vehicle connectivities (a) and distance thresholds (b)
  • Figure 4: Illustration of incorrect predictions. Observations (green), blue (ground truth), and predictions (red) are shown.