Incremental Semantic Mapping with Unsupervised On-line Learning

  • 2019-07-11 03:52:13
  • Ygor C. N. Sousa, Hansenclever F. Bassani
  • 0

Abstract

This paper introduces an incremental semantic mapping approach, with on-lineunsupervised learning, based on Self-Organizing Maps (SOM) for robotic agents.The method includes a mapping module, which incrementally creates a topologicalmap of the environment, enriched with objects recognized around eachtopological node, and a module of places categorization, endowed with anincremental unsupervised learning SOM with on-line training. The proposedapproach was tested in experiments with real-world data, in which itdemonstrates promising capabilities of incremental acquisition of topologicalmaps enriched with semantic information, and for clustering together similarplaces based on this information. The approach was also able to continuelearning from newly visited environments without degrading the informationpreviously learned.

 

Quick Read (beta)

loading the full paper ...