Machine Learning to get Poppy walking


#1

I’m new. So new that this is my first post! So, in the spirit of community: Hello!

Now, ever since I found the poppy project I have wanted to get it walking. While there are probably more than a few of you working on this sort of thing I have yet to find a someone who has implemented a natural, adaptive gait.

I haven’t been here long, but using the search function I have found little discussion on the topic of machine learning here. Perhaps I am mistaken?

Has anyone tried a machine learning solution? Not just on V-Rep but also on their working poppy? If you have, I would enjoy seeing your method and results.

Cheers!


#2

Hello to you and welcome!

Indeed there has been little development around learning to walk since the thesis of @Matthieu . But the Learning remains a central theme in the Inria - FLOWERS team:

«The Flowers team studies computational mechanisms allowing robots and humans to acquire open-ended repertoires of skills through life-long learning. This includes the processes for progressively discovering their bodies and interaction with objects, tools and others. In particular, we study mechanisms of intrinsically motivated learning (also called curiosity-driven active learning), autonomous unsupervised exploration, imitation and social learning, multimodal statistical inference, embodiment and maturation and self-organization.» https://www.inria.fr/en/teams/flowers

Maybe you can check the team forum: https://openlab-flowers.inria.fr/