Home Research Teaching Projects Publications

Short info:

I am a researcher with the Flowers' team at INRIA. I am interested in the use of machine learning, control theory to create robots that are more robust and can interact and work with people. My research interests are on learning by imitation, artificial development and active learning approaches.

News:

myphoto
Manuel Lopes
FLOWERS team
INRIA Bordeaux Sud-Ouest, Bat. A29
351 Cours de la Libération
33405 Talence, Cedex
France

manuel . lopes @ inria . fr

Phone: +33524574179

Marie Sanchez (administrative assistant) + 33 5 40 00 38 24



Slideshow image


Since your web browser does not support JavaScript, here is a non-JavaScript version of the image slideshow:

slideshow image
This images, inspired on Gibson's visual space, shows that the same gesture can have an huge difference in appearance. When doing imitation this visual perspective must be taken into account.

Image idea Manuel Lopes, Design Francisco Melo


slideshow image
Baltazar grasps an object. Grasping is a very complex skill, it requires localizing the object, choose a good grasping point, plan a reaching behavior, execute the reaching and final close the hand.


slideshow image
Baltazar likes to play with objects. In this case it is executing a task that was learned from a demonstration.


slideshow image
Baltazar likes carrots!!!


slideshow image
Robots are very complex machines. This is the team that design and built it.


slideshow image
iCub can localize sounds. This is how it does it.


slideshow image
Baltazar likes to imitate. Here it copies only the observed effects without taking into account the demonstrator.


slideshow image
iCub likes to know everything that goes around him. An egosphere includes information about the motion, color, sound and edges that are around.


slideshow image
When imitating we do not consider everything. When saying goodbye do you use the same elbow motion? Or just the hand motion?


slideshow image
How can you learn a relation where a single input can have more than one output?


slideshow image
Policy gradient methods are a good choice to improve control policies. How can you reuse all your data?


slideshow image
Gesture imitation is easy. Or maybe not...


slideshow image
One of the first pictures of the iCub head.


slideshow image
One of the first pictures of the iCub head before being built!


slideshow image
iCub is tested as a real child.


slideshow image
iCub with very nice ears. Symmetrical ears are bad!


slideshow image
Sometimes we need a good wizard of Oz.


slideshow image
Imitation is not just for learning. It is also for social acceptance.


slideshow image
A mosaic of the space around iCub.


slideshow image
What is the stationary point of two learners that do not know that they are playing a game?


slideshow image
A parallel mechanism was also studied as a potential iCub neck.


slideshow image
Baltazar showing its skills.


slideshow image
Baltazar design phase.