direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Inhalt des Dokuments

Predicting Kinematic Degrees of Freedom from RGB-D Data

Lupe

Motivation

A large part of our everyday environment consist of rigid bodies and kinematic joints that connect them and restrict their relative motion. Together, these joints and bodies form kinematic structures called articulated objects, such as cupboards, drawers, doors etc. Robots that aim to work in our homes or offices need to be able to explore, analyze and manipulate such kinematic structures.

 

An important aspect of this exploration and the manipulation of new articulated objects is the generation of initial hypotheses about the type, location and other parameters of kinematic joints, based on visual input. If a robot can generate a good set of initial hypotheses, it could explore more efficiently, reducing the number of actions that it needs to reveal kinematic degrees of freedom. In this thesis you will develop a method for predicting kinematic degrees of freedom from RGB-D data.

 

Description of work

You will work on a module that takes RGB-D images / videos as input and predicts kinematic joints. You will use neural networks and train them on a dataset that contains RGB-D images of furniture, labeled with ground truth about prismatic and rotational degrees of freedom.

 

Requirements

Strong C++ and/or python programming skills

Computer Vision

Machine Learning

Robotics (ROS)

 

Contact

Manuel Baum

Roberto Martín-Martín

Oliver Brock

back

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions