Improved and scalable online learning of spatial concepts and language models with mapping

We propose a novel online learning algorithm, called SpCoSLAM 2.0, for spatial concepts and lexical acquisition with high accuracy and scalability. Previously, we proposed SpCoSLAM as an online learning algorithm based on unsupervised Bayesian probabilistic model that integrates multimodal place cat...

Full description

Saved in:
Bibliographic Details
Published inAutonomous robots Vol. 44; no. 6; pp. 927 - 946
Main Authors Taniguchi, Akira, Hagiwara, Yoshinobu, Taniguchi, Tadahiro, Inamura, Tetsunari
Format Journal Article
LanguageEnglish
Published New York Springer US 01.07.2020
Springer Nature B.V
Subjects
Accuracy
Algorithms
Artificial Intelligence
Computer Imaging
Control
Distance learning
Engineering
Machine learning
Mapping
Mechatronics
Pattern Recognition and Graphics
Probabilistic models
Robotics
Robotics and Automation
Training
Vision
Place categorization
Semantic mapping
Scalability
Online learning
Lexical acquisition
Unsupervised Bayesian probabilistic model
Online AccessGet full text

Cover

More Information
Summary:We propose a novel online learning algorithm, called SpCoSLAM 2.0, for spatial concepts and lexical acquisition with high accuracy and scalability. Previously, we proposed SpCoSLAM as an online learning algorithm based on unsupervised Bayesian probabilistic model that integrates multimodal place categorization, lexical acquisition, and SLAM. However, our original algorithm had limited estimation accuracy owing to the influence of the early stages of learning, and increased computational complexity with added training data. Therefore, we introduce techniques such as fixed-lag rejuvenation to reduce the calculation time while maintaining an accuracy higher than that of the original algorithm. The results show that, in terms of estimation accuracy, the proposed algorithm exceeds the original algorithm and is comparable to batch learning. In addition, the calculation time of the proposed algorithm does not depend on the amount of training data and becomes constant for each step of the scalable algorithm. Our approach will contribute to the realization of long-term spatial language interactions between humans and robots.
ISSN:0929-5593
1573-7527
DOI:10.1007/s10514-020-09905-0