New Music Artist Videos
20 Videos. Showing Newest from #1
Embed to Control: A Locally Linear Latent Dynamics Model for Control from Raw Images

Add to EJ Playlist  We introduce Embed to Control (E2C), a method for model learning and control of non-linear dynamical systems from raw pixel images. E2C consists of a deep generative model, belonging to the family of variational autoencoders, that learns to generate image trajectories from a latent space in which the dynamics is constrained to be locally linear. Our model is derived directly from an optimal control formulation in latent space, supports long-term prediction of image sequences and exhibits strong performance on a variety of complex control problems. The Paper can be found on arxiv: http://arxiv.or g/abs/1506.0736 5

Approximate Real-Time Optimal Control Based on Sparse Gaussian Process Models

Add to EJ Playlist  a.k.a. Learning to Swing-Up and Balance from Scratch in under 3 Minutes Revisiting the classical cart-pole balancing system (https://www.yo ?v=Lt-KLtkDlh8) , we present a fully automated system for (approximate) optimal control of non-linear systems. Our approach jointly learns a non-parametric model of the system dynamics and performs receding horizon control. This results in an extremely data-efficient learning algorithm that can operate under real-time constraints. Our algorithm successfully learns to control a real cart-pole system from-scratch (without prior knowledge provided by an expert), in less than 10 episodes of interaction with the system, amounting to less than 3 minutes of real time. For details, see: Joschka Boedecker, Jost Tobias Springenberg, Jan W├╝lfing, Martin Riedmiller (2014) Approximate Real-Time Optimal Control Based on Sparse Gaussian Process Models. In Adaptive Dynamic Programming and Reinforcement Learning (ADPRL).

BCI for High-Level Remote Reaching and Grasping

Add to EJ Playlist  Showcase of the prototype system developed during the first stage of the NeuroBots project ( ainlinks-braint ools.uni-freibu projects/projec t-details/neuro bots) as a collaboration between the iEEG Lab (http://www.iee g.uni-freiburg. de) and the Machine Learning Lab ( rmatik.uni-frei of the University of Freiburg. Imagined motor commands are used for high-level remote control of an autonomous, reinforcement-l earning-based robotic system for reaching and grasping several kilometers away.

Combining EEG and Reinforcement Learning for Robot Control

Add to EJ Playlist  Teaser video for the BrainLinks-Brai nTools Cluster of Excellence ( ainlinks-braint ools.uni-freibu Improving on our previous system ( v=ZIQ6uyptAB8), we use electroencephal ography to communicate higher-level intentions to an autonomous controller learned via Neural Fitted Q-Iteration.

Neural Controller for Ms. Pac-Man

Add to EJ Playlist  Learning a controller for the Ms. Pac-Man arcade game by trial and failure. The learning algorithm is TD lambda using a neural network for value function representation and advanced features as input.

A Visual Servoing Approach to Manipulation using Neural Reinforcement Learning

Add to EJ Playlist  A Kinova Jaco robotic arm is expected to learn how to reach for an object and pick it up autonomously. Control is realized in a visual feedback control loop, making it both reactive and robust to noise. The controller is learned from scratch, without prior knowledge of proper behaviour, by success or failure using Neural Fitted Q Iteration.

Visual Deep Learning

Add to EJ Playlist  Swinging Up and Balancing a Real Pole

Carrera Vision Tracking

Add to EJ Playlist  Summary of the bachelor thesis project by Manuel Watter. Using Neural Fitted Q-Iteration, a robot head first learns to keep a stationary object centered in the camera, and then to adjust its actions so as to better keep it in view once it is moving.

Combining EOG and NFQ for Robot Control

Add to EJ Playlist  First attempt at combining electrooculogra phy and machine learning to have a Katana robotic arm search for an object and pick it up. Collaboration between the Machine Learning Lab and the Biomedical Microsystems Group.

Learning to Dribble by Success and Failure

Add to EJ Playlist  The behavior for dribbling the ball is trained by reinforcement learning


Add to EJ Playlist  reinforcement learning slotcar system

Harting Workshop

Add to EJ Playlist  Programming workshop with apprentices of Harting

Evolution of the Brainstormers Tribots

Add to EJ Playlist  This video summarizes the development process of the Tribots robot soccer team from the first experiments in a student project until winning the world championships 2006 and 2007.

RoboCup 2007 Atlanta, Brainstormers Tribots Mix

Add to EJ Playlist  Best scenes from the RoboCup 2007 tournament in Atlanta.

Passing Soccer Robots

Add to EJ Playlist  Pass play in robot soccer. Actually, this was the very first completed pass of the Brainstormers Tribots in an official tournament.

NEWS   Top Videos   Music   Classical   Listen   Funny   Fails   Artist List   Aww   Gaming   Minisode   Science   Technology   TED   TWiT   Trailers   M.M.   PBS   WSJ   AP   CSPAN   CNN   RT   TMZ   E!   ABC   CBS   Politics   Sports   ESPN   WTF   Conspiracy  

Related Uploaders:

tribots Playlists:
Reinforcement Learning for Games