Allo-centric Occupancy Grid Prediction for Urban Traffic Scene Using Video Prediction Networks

TL;DR

This paper introduces an allo-centric occupancy grid framework for urban traffic scene prediction, improving long-term forecasting accuracy over traditional egocentric methods by maintaining static scene structure.

Contribution

The paper proposes a novel allo-centric occupancy grid representation for traffic scene prediction, enabling better long-term forecasts and scene structure preservation in autonomous vehicle navigation.

Findings

Allo-centric grid significantly outperforms ego-centric grid in prediction accuracy.

Static scene representation helps maintain scene structure at turns.

Approach validated on real-world Nuscenes dataset.

Abstract

Prediction of dynamic environment is crucial to safe navigation of an autonomous vehicle. Urban traffic scenes are particularly challenging to forecast due to complex interactions between various dynamic agents, such as vehicles and vulnerable road users. Previous approaches have used egocentric occupancy grid maps to represent and predict dynamic environments. However, these predictions suffer from blurriness, loss of scene structure at turns, and vanishing of agents over longer prediction horizon. In this work, we propose a novel framework to make long-term predictions by representing the traffic scene in a fixed frame, referred as allo-centric occupancy grid. This allows for the static scene to remain fixed and to represent motion of the ego-vehicle on the grid like other agents'. We study the allo-centric grid prediction with different video prediction networks and validate the approach on the real-world Nuscenes dataset. The results demonstrate that the allo-centric grid representation significantly improves scene prediction, in comparison to the conventional ego-centric grid approach.