Explauto: an open-source Python library to study autonomous exploration in developmental robotics

Clément Moulin-Frier 1, * Pierre Rouanet 1 Pierre-Yves Oudeyer 1
* Corresponding author
1 Flowers - Flowing Epigenetic Robots and Systems
Inria Bordeaux - Sud-Ouest, U2IS - Unité d'Informatique et d'Ingénierie des Systèmes
Abstract : We present an open-source Python library, called Explauto, providing a unified API to design and compare various exploration strategies driving various sensorimotor learning algorithms in various simulated or robotics systems. Explauto aims at being collaborative and pedagogic, providing a platform to developmental roboticists where they can publish and compare their algorithmic contributions related to autonomous exploration and learning, as well as a platform for teaching and scientific diffusion. The library is available at this address: https://github.com/flowersteam/explauto
Keywords : teaching Open-source software Exploration Active Learning Goal babbling Intrinsic Motivation teaching.
Document type :
Conference papers
ICDL-Epirob - International Conference on Development and Learning, Epirob, Oct 2014, Genoa, Italy. 2014
Liste complète des métadonnées

Cited literature [8 references]  Display  Hide  Download
https://hal.inria.fr/hal-01061708
Contributor : Clément Moulin-Frier <>
Submitted on : Monday, September 8, 2014 - 12:25:35 PM
Last modification on : Thursday, April 12, 2018 - 1:06:49 PM
Document(s) archivé(s) le : Tuesday, December 9, 2014 - 11:50:22 AM

File

icdl2014_cmf_pr_pyo.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01061708, version 1

Collections

INRIA | ENSTA | PARISTECH

Citation

Clément Moulin-Frier, Pierre Rouanet, Pierre-Yves Oudeyer. Explauto: an open-source Python library to study autonomous exploration in developmental robotics. ICDL-Epirob - International Conference on Development and Learning, Epirob, Oct 2014, Genoa, Italy. 2014. 〈hal-01061708〉

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

1072

Files downloads

580

Contact


archive-ouverte@inria.fr