no code implementations • 24 Jun 2021 • Ygor C. N. Sousa, Hansenclever F. Bassani
In contrast, this work introduces a topological semantic mapping method that uses deep visual features extracted by a CNN (GoogLeNet), from 2D images captured in multiple views of the environment as the robot operates, to create, through averages, consolidated representations of the visual features acquired in the regions covered by each topological node.
no code implementations • 9 Jul 2019 • Ygor C. N. Sousa, Hansenclever F. Bassani
This paper introduces an incremental semantic mapping approach, with on-line unsupervised learning, based on Self-Organizing Maps (SOM) for robotic agents.