Occupancy Grid Mapping with Cognitive Plausibility for Autonomous Driving Applications

Abstract

This work investigates the validity of an occupancy grid mapping inspired by human cognition and the way humans visually perceive the environment. This query is motivated by the fact that, to date, no autonomous driving system reaches the performance of an ordinary human driver. The mechanisms behind human perception could provide cues on how to improve common techniques employed in autonomous navigation-specifically the use of occupancy grids to represent the environment. We experiment with a neural network that maps an image of the scene onto an occupancy grid representation, and we show how the model benefits from two key (and yet simple) changes: 1) a different format of occupancy grid that resembles the way the brain projects the environment into a warped representation in the cortical visual area; 2) a mechanism similar to human visual attention that filters out non-relevant information from the scene. These effective expedients can potentially be applied to any autonomous driving task requiring an abstract representation of the scenario like the occupancy grids.

Distributed cognition for collaboration between human drivers and self-driving cars

“…Therefore, there is no need anymore for the user to drive for a moment to the center to have a clearer view of the road, because the agent is the one responsible to check if the overtake is feasible. Even if the user steers to the side to see ahead, the agent will not execute the overtake if it deems the maneuver risky (note that the perception system of the agent differs from the human, so the agent does not actually need to drive to the side to see better, like a human driver would do—although there are new attempts at human-inspired perception for autonomous vehicles; Plebe et al, 2021 ). This collaboration paradigm leads to a new way of driving, and human drivers will need some time to adjust to it.…”

HMM-Based Dynamic Mapping with Gaussian Random Fields

“…Instead of observation models, posterior probability can be used directly to estimate HMM parameters, which is convenient in computation in this paper. (2) The HMM parameters of observed space can be smoothed. GRFs are applied to consider the dependence between HMM parameters of different points.…”

“…In earlier research, the environments were assumed to be static. The classical method for static environments is occupancy grid mapping [1][2][3] where maps are divided into a grid and the states of different grid cells are assumed to be independent. In dynamic environments, one popular strategy is to estimate the number of potential targets, their positions, and velocities from sensor data [4][5][6].…”

Amodal Cityscapes: A New Dataset, its Generation, and an Amodal Semantic Segmentation Challenge Baseline

“…This yields essential visual cues for scene understanding, and thus forms a significant part of environment perception. One could argue that amodal perception should be performed after fusion in the occupancy grid/vector space [10], [11], [12], or in some latent space non-interpretable to humans [13], [14]. This work on amodal perception of simple camera data, however, comes with the clear advantage of (a) being human-readable, and (b) that already during multisensor fusion, the redundancy of multiple estimates of occluded objects may yield powerful confidence information.…”