Home
People
Projects
Research
Publications
Surveys
Courses
Student Projects
Jobs
Downloads
This is an old revision of the document!
Making robots learn to perceive and act with understanding
At IIS we enable autonomous robots to perceive and act flexibly and robustly in unstructured environments, leveraging machine learning methods to build perceptual, motor and reasoning skills.
We seek to answer the question: How can we enable robots to acquire the knowledge and understanding they require to interact sensibly with unstructured environments?
Our research addresses complete perception-action loops, from computer vision to grasping and manipulation, using reactive algorithms and/or cognitive models. Much of our work uses machine learning to enable robots to synthesize and improve complex and robust sensorimotor behavior with experience. Related areas of interest include human-robot interaction, image and video analysis, and visual neuroscience.
2021-03-24 | Justus Piater gives an invited lecture Machine Learning, Perception, And Abstract Concepts at Invited lecture at Ontario Tech U, Canada, online. [Abstract]With every spectacular achievement of a machine learning system, the long-elusive AI breakthrough is popularly proclaimed to be just around the corner. Most recent successes have been due in large part to massive data and computation, in particular using deep artificial neural networks. But can artificial cognition really be achieved just by further scaling up existing machine-learning techniques? I discuss examples of simple, perceptual problems that are easily solved by humans but very difficult for today's machine learning methods. These problems reflect how humans conceptualize their world. Their mastery is thus likely to be an essential prerequisite for autonomous robots to attain higher levels of cognitive abilities. To get there, a few core issues can be identified that should drive research in cognitive robotics. |
2021-01-28 | Matteo Saveriano gives an invited talk Making robots to learn from human observation? at Showcasing Young Austrian Scholars and Scientists, Austrian cultural forum, Ottawa, online. |
2021-01-14 | Matteo Saveriano gives an invited talk Hierarchical action decomposition and motion learning for the execution of manipulation tasks at Hello Tyrol calling! Robotics Talk online, GMAR, Innsbruck, online. |
2020-11-19 | Justus Piater gives an invited talk Machine Learning in Robotics at bAIome PI Talk, Center for
Biomedical AI, University Medical Center Hamburg-Eppendorf, online. [Abstract]Machine Learning increasingly equips robots with learning capabilities and flexibility. This will enable them to act purposefully in unstructured environments and to react to unforeseen events. People can teach them intuitively to perform tasks instead of having to program them in painstaking ways. Robots can learn from experience and can improve their behavior over time. In this talk I will give an overview of methods, opportunities, and challenges of machine learning in robotics. |
2020-11-20 | Simon Haller-Seeber and Patrick Lamprecht present a show Explainable AI: A sneak peek into the Black-Box at Science Slam, online. |
2020-10-22 – 2020-10-23 | Matteo Saveriano, Erwan Renaudo, Antonio Rodríguez-Sánchez, and Justus Piater organize the 13th International Workshop on Human-Friendly Robotics (HFR 2020), Innsbruck (online). |
2020-09-30 | Erwan Renaudo contributes a talk ROSSINI: RobOt kidS deSIgn thiNkIng at Robotics in Education 2020, online. |
2020-06-22 | Justus Piater gives an invited talk Conditional Neural Movement Primitives at GdR
ISIS Réunion Apprentissage et Robotique, online. [Abstract]Conditional Neural Movement Primitives (CNMP) constitute a novel framework for robot programming by demonstration based on Conditional Neural Processes (CNP). Like Bayesian methods such as Gaussian Processes (GP), CNP learn how target distributions depend on data, and can be conditioned on specific data points to infer new target distributions at test time. Unlike GP that are expensive to train and scale poorly to high dimensions, CNP are neural networks and are trained by gradient descent. CNMP leverage CNP to represent motion trajectories that can be conditioned, at test time, on task paramters such as goal locations, via points, and/or force readings. Moreover, CNMP are conditioned on sensor readings during execution, resulting in robust, reactive behavior. This talk will present an overview of how CNMP work and how they can be used in various robot applications. |
2020-06-03 | Justus Piater appears in the media: Wie der Roboter denken lernt. |
2020-01-29 | Justus Piater gives an invited talk Digital Science at Vortragsreihe
„Primers for Predocs – Strategien für eine erfolgreiche
Promotion“, Universität Heidelberg. [Abstract]Massive availability of data and computing power are promoting data-driven methods in all areas of science and technology. I will describe how the University of Innsbruck supports this via its new Digital Science Center, and will give a flavor of machine learning for data analysis. |
2020-01-20 | Joanna Chimiak-Opoka, Carina König, and Justus Piater appear in the media: Ergänzung Digital Science erfolgreich gestartet – UIBK Newsroom. |
University of Innsbruck
Department of Computer Science
Technikerstr. 21a
6020 Innsbruck
Austria
How to find us: See the directions.
Legal Notice: See the Impress and Privacy Notice.