.png){kind=icon}
.png){kind=icon}
Return
A Multisensor Fusion Method Based on High-Dimensional Cognitive Encoding for Large-Scale Navigation
.png){kind=icon}
PRE
AI
DOI：10.1109/TIM.2025.3585212.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
With the widespread application of unmanned systems in large-scale and complex environments, accurate and reliable navigation has become a pivotal challenge. Most existing brain-inspired multisensor fusion methods primarily focus on small-scale environments and overlook the impact of high-dimensional cognitive information on navigation. In this article, inspired by the navigation mechanisms of the bat brain, we propose a multisensor fusion navigation architecture based on memory and reasoning mechanisms. The proposed architecture adaptively integrates multisensor information in high-dimensional cognitive space through environmental perception, spatial perception, and cognitive reasoning. To further improve navigation performance, we propose a place cell (PC) update method based on cyclic mapping and a cognitive map decoding method based on interest iteration. The proposed method is evaluated on both public datasets and our own dataset. Experimental results demonstrate that the proposed brain-inspired multisensor fusion method outperforms existing methods in terms of localization accuracy and robustness, offering a promising brain-inspired solution for large-scale navigation tasks.
Keywords:
Brain-inspired navigation
cognitive map
cognitive space
continuous attractor network
large-scale environment
multisensor fusion
Journal
IF:
5.9
Papers:
1.9W
Citations:
5.8W