Search Results for author: Jens Kober

Found 12 papers, 5 papers with code

ILoSA: Interactive Learning of Stiffness and Attractors

no code implementations4 Mar 2021 Giovanni Franzese, Anna Mészáros, Luka Peternel, Jens Kober

Teaching robots how to apply forces according to our preferences is still an open challenge that has to be tackled from multiple engineering perspectives.

Gaussian Processes

GEM: Glare or Gloom, I Can Still See You -- End-to-End Multimodal Object Detection

no code implementations24 Feb 2021 Osama Mazhar, Robert Babuska, Jens Kober

We additionally record a new RGB-Infra indoor dataset, namely L515-Indoors, and demonstrate that the proposed object detection methodologies are highly effective for a variety of lighting conditions.

2D Object Detection Human robot interaction

Random Shadows and Highlights: A new data augmentation method for extreme lighting conditions

1 code implementation13 Jan 2021 Osama Mazhar, Jens Kober

In this paper, we propose a new data augmentation method, Random Shadows and Highlights (RSH) to acquire robustness against lighting perturbations.

Data Augmentation

DeepKoCo: Efficient latent planning with a robust Koopman representation

no code implementations25 Nov 2020 Bas van der Heijden, Laura Ferranti, Jens Kober, Robert Babuska

This paper presents DeepKoCo, a novel model-based agent that learns a latent Koopman representation from images.

Interactive Imitation Learning in State-Space

1 code implementation2 Aug 2020 Snehal Jauhri, Carlos Celemin, Jens Kober

Imitation Learning techniques enable programming the behavior of agents through demonstrations rather than manual engineering.

Imitation Learning

Smooth Exploration for Robotic Reinforcement Learning

4 code implementations12 May 2020 Antonin Raffin, Jens Kober, Freek Stulp

We evaluate gSDE both in simulation, on PyBullet continuous control tasks, and directly on three different real robots: a tendon-driven elastic robot, a quadruped and an RC car.

Continuous Control

Deep Reinforcement Learning with Feedback-based Exploration

2 code implementations14 Mar 2019 Jan Scholten, Daan Wout, Carlos Celemin, Jens Kober

We employ binary corrective feedback as a general and intuitive manner to incorporate human intuition and domain knowledge in model-free machine learning.

Continuous Control OpenAI Gym

Learning Gaussian Policies from Corrective Human Feedback

no code implementations12 Mar 2019 Daan Wout, Jan Scholten, Carlos Celemin, Jens Kober

We demonstrate that the novel algorithm outperforms the current state-of-the-art in final performance, convergence rate and robustness to erroneous feedback in OpenAI Gym continuous control benchmarks, both for simulated and real human teachers.

Continuous Control Gaussian Processes +1

Interactive Learning with Corrective Feedback for Policies based on Deep Neural Networks

1 code implementation30 Sep 2018 Rodrigo Pérez-Dattari, Carlos Celemin, Javier Ruiz-del-Solar, Jens Kober

Deep Reinforcement Learning (DRL) has become a powerful strategy to solve complex decision making problems based on Deep Neural Networks (DNNs).

Car Racing Decision Making

Using Bayesian Dynamical Systems for Motion Template Libraries

no code implementations NeurIPS 2008 Silvia Chiappa, Jens Kober, Jan R. Peters

Motor primitives or motion templates have become an important concept for both modeling human motor control as well as generating robot behaviors using imitation learning.

Imitation Learning Time Series

Policy Search for Motor Primitives in Robotics

no code implementations NeurIPS 2008 Jens Kober, Jan R. Peters

We compare this algorithm to alternative parametrized policy search methods and show that it outperforms previous methods.

Imitation Learning Policy Gradient Methods

Cannot find the paper you are looking for? You can Submit a new open access paper.