Autonomous Knowledge Discovery Based on Artificial Curiosity-Driven Learning by Interaction

K. Madani 1 D. M. Ramik 1 C. Sabourin 1
1 SYNAPSE
LISSI - Laboratoire Images, Signaux et Systèmes Intelligents
Abstract : In this work, we investigate the development of a real-time intelligent systemallowing a robot to discover its surrounding world and to learn autonomouslynew knowledge about it by semantically interacting with humans. The learningis performed by observation and by interaction with a human. We describe thesystem in a general manner, and then we apply it to autonomous learning ofobjects and their colors. We provide experimental results both using simulatedenvironments and implementing the approach on a humanoid robot in a real-world environment including every-day objects. We show that our approachallows a humanoid robot to learn without negative input and from a smallnumber of samples.
Type de document :
Chapitre d'ouvrage
R. Duro and Y. Kondratenko. Advances in Intelligent Robotics and Collaborative Automation, River Publishers, pp.73-94, 2015
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01682086
Contributeur : Lab Lissi <>
Soumis le : jeudi 11 janvier 2018 - 23:47:44
Dernière modification le : lundi 18 février 2019 - 17:02:02

Identifiants

  • HAL Id : hal-01682086, version 1

Collections

Citation

K. Madani, D. M. Ramik, C. Sabourin. Autonomous Knowledge Discovery Based on Artificial Curiosity-Driven Learning by Interaction. R. Duro and Y. Kondratenko. Advances in Intelligent Robotics and Collaborative Automation, River Publishers, pp.73-94, 2015. 〈hal-01682086〉

Partager

Métriques

Consultations de la notice

47