Differences

This shows you the differences between two versions of the page.

--- start [2020/11/26 06:25]
IIS Webadmin
+++ start [2022/06/09 06:25]
IIS Webadmin
@@ Line 11: / Line 11: @@
 ===== Working With Us =====
+  * We are hiring [[jobs|two doctoral students]] in machine learning for computer vision or robotics.
   * Check our thesis topics for [[theses/start|Bachelor and Master students]].
-  * [[https://ellis.eu/|ELLIS]] PhD Program: [[https://ellis.eu/de/news/ellis-phd-program-call-for-applications|Call for Applications]]
+ <html><!--
+ * [[https://ellis.eu/|ELLIS]] PhD Program: [[https://ellis.eu/news/ellis-phd-program-call-for-applications-deadline-november-15-2021|Call for Applications]]
+--></html>
 <html><!--
@@ Line 29: / Line 32: @@
 ===== News =====
-<HTML><table><tr valign="top"><td>2020-11-19</td><td>Justus Piater gives an invited talk <i>Machine Learning in Robotics</i> at <a href="https://baiome.org/">bAIome PI Talk, Center for
+<HTML><table><tr valign="top"><td>2022-06-02</td><td>Erwan Renaudo contributes a talk <i>Deep Learning for Fast Segmentation of E-waste Devices’
-    Biomedical AI, University Medical Center Hamburg-Eppendorf</a>, online. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem1', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem1Abstract" style="display:none">Machine Learning increasingly equips robots with
+    Inner Parts in a Recycling Scenario</i> at <a href="https://icprai2022.sciencesconf.org/">ICPRAI 2022</a>, Paris/online. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem0', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem0Abstract" style="display:none">Recycling obsolete electronic devices (E-waste) is a
-    learning capabilities and flexibility.  This will enable them to
+    dangerous task for human workers. Automated E-waste recycling
-    act purposefully in unstructured environments and to react to
+    is an area of great interest but challenging for current robotic
-    unforeseen events.  People can teach them intuitively to perform
+    applications. We focus on the problem of segmenting inner parts
-    tasks instead of having to program them in painstaking ways.
+    of E-waste devices into manipulable elements. First, we extend
-    Robots can learn from experience and can improve their behavior
+    a dataset of hard-drive disk (HDD) components with labelled
-    over time.  In this talk I will give an overview of methods,
+    occluded and non-occluded points of view of the parts, in order
-    opportunities, and challenges of machine learning in
+    to increase the diversity and the quality of the learning data
-    robotics.</blockquote></td></tr><tr valign="top"><td>2020-11-20</td><td>Simon Haller-Seeber and Patrick Lamprecht present a show <i>Explainable AI: A sneak peek into the Black-Box</i> at <a href="https://youtube.com/watch?v=4NVfPwdLnCg&amp;t=17m43s">Science Slam</a>, online.</td></tr><tr valign="top"><td>2020-06-22</td><td>Justus Piater gives an invited talk <i>Conditional Neural Movement Primitives</i> at <a href="http://www.gdr-isis.fr/index.php?page=reunion&amp;idreunion=424">GdR
+    with different angles. We then perform an extensive evaluation
-    ISIS Réunion Apprentissage et Robotique</a>, online. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem3', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem3Abstract" style="display:none">Conditional Neural Movement Primitives (CNMP) constitute
+    with three different state-of-the-art models, namely CenterMask,
-    a novel framework for robot programming by demonstration based on
+    BlendMask and SOLOv2 (including variants) and two types of
-    Conditional Neural Processes (CNP).  Like Bayesian methods such as
+    metrics: the average precision as well as the frame rate. Our
-    Gaussian Processes (GP), CNP learn how target distributions depend
+    results show that instance segmentation using state-of-the-art
-    on data, and can be conditioned on specific data points to infer
+    deep learning methods can precisely detect complex shapes along
-    new target distributions at test time.  Unlike GP that are
+    with their boundaries, as well as being suited for fast tracking of
-    expensive to train and scale poorly to high dimensions, CNP are
+    parts in a robotic recycling system. (Rojas et  al. 2022)
-    neural networks and are trained by gradient descent.  CNMP
+    </blockquote></td></tr><tr valign="top"><td>2022-05-20</td><td>Justus Piater appears in the media: <a href="https://radiothek.orf.at/podcasts/tir/derwoche/derwoche-maschinenzukunft">ORF
-    leverage CNP to represent motion trajectories that can be
+    Tirol Podcast (in German)</a>. (Künstliche Intelligenz (KI) ist bereits heute ein täglicher Begleiter, doch vom universellen Haushaltsroboter sind wir noch weit entfernt. Automatisierte Systeme werden zwar Jobs übernehmen, Maschinen werden aber noch lange nicht die Weltherrschaft an sich reißen, falls überhaupt. Zur Langen Nacht der Forschung spricht Martin Kubin im #derWoche „Maschinenzukunft“ mit Prof. Justus Piater vom Institut für Informatik über Chancen und Gefahren von KI. Er weiß auch, warum KI etwa die Justiz unterstützen kann, aber ein KI-Richter niemals objektiv urteilen könnte.)</td></tr><tr valign="top"><td>2022-05-17</td><td>Erwan Renaudo gives an invited talk <i>A Brief tour of autonomous robots' &quot;Cognition&quot;</i> at <a href="https://www.philosophie.uni-konstanz.de/ag-mueller/aktuelles/meldungsdetails-ag-mueller/2022/5/17/event/46426-Creating-Agency-and-Cogn/tx_cal_phpicalendar/">Creating Agency and Cognition in Automated Systems: What can we learn from the Octopus? International Hybrid Workshop</a>, Innsbruck. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem2', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem2Abstract" style="display:none">In the design of &quot;Cognitive Robots&quot; software architectures,
-    conditioned, at test time, on task paramters such as goal
+    researchers had to face many organizational problems due to
-    locations, via points, and/or force readings.  Moreover, CNMP are
+    the particular nature of robots. Not only should they manage
-    conditioned on sensor readings during execution, resulting in
+    to achieve complex tasks with long-terms goals, but they must
-    robust, reactive behavior.  This talk will present an overview of
+    also be reactive to their surroundings to preserve their environment.
-    how CNMP work and how they can be used in various robot
+    We will go over several layered control paradigms developed over
-    applications.</blockquote></td></tr><tr valign="top"><td>2020-06-03</td><td>Justus Piater appears in the media: Wie der Roboter denken lernt.</td></tr><tr valign="top"><td>2020-01-29</td><td>Justus Piater gives an invited talk <i>Digital Science</i> at <a href="https://www.graduateacademy.uni-heidelberg.de/karriere/veranstaltungsreihen.html">Vortragsreihe
+    the history of robotics and see how these can relate to the
-    „Primers for Predocs – Strategien für eine erfolgreiche
+    knowledge about the Octopus' brain organization, and discuss
-    Promotion“</a>, Universität Heidelberg. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem6', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem6Abstract" style="display:none">Massive availability of data and computing power are
+    how the Octopus itself can inspire cognitive robotic research.
-    promoting data-driven methods in all areas of science and
+    </blockquote></td></tr><tr valign="top"><td>2022-05-11</td><td>Justus Piater gives an invited talk <i>Robots That Learn Like Humans?</i> at <a href="https://pintofscience.at/events/innsbruck">Pint of
-    technology.  I will describe how the University of Innsbruck
+    Science</a>, Innsbruck. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem3', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem3Abstract" style="display:none">Wouldn't it be great to have a helper robot that we can
-    supports this via its new Digital Science Center, and will give a
+    task with our annoying chores? How would we teach the robot to
-    flavor of machine learning for data analysis.</blockquote></td></tr><tr valign="top"><td>2020-01-20</td><td>Joanna Chimiak-Opoka, Carina König, and Justus Piater appear in the media: <a href="https://www.uibk.ac.at/newsroom/ergaenzung-digital-science-erfolgreich-gestartet.html.de">Ergänzung Digital Science erfolgreich gestartet – UIBK Newsroom</a>.</td></tr><tr valign="top"><td>2020-01-03</td><td>Justus Piater gives an invited talk <i>Künstliche Intelligenz: Grundlagen, Erfolge,
+    perform these tasks? Can we build such robots by equipping them
-    Herausforderungen</i> at 47. Tagung des Innsbrucker Kreises von MoraltheologInnen
+    with artificial intelligence? We will discuss how human and
-    und SozialethikerInnen, Innsbruck.</td></tr><tr valign="top"><td>2019-12-19</td><td>Justus Piater appears in the media: <a href="https://tvthek.orf.at/profile/Tirol-heute/70023/Tirol-heute/14035670/Alexa-Siri-Co-Spione-im-eigenen-Haus/14610743">TV interview by ORF 2 Tirol Heute RedHaus (in German)</a>.</td></tr><tr valign="top"><td>2019-12-12</td><td>Justus Piater gives an invited lecture <i>Too Smart to Be Trusted – Do I Even Want to
+    machine learning differ and how this impacts what skills,
-    Understand My Robot?</i> at <a href="http://www.trustrobots.eu/">TrustRobots Lecture series
+    including for communication, our helper robot will be able to
-    Trust in Robots</a>, TU Vienna.</td></tr><tr valign="top"><td>2019-12-05</td><td>IIS guest <span style="font-weight:bold">Heiko Neumann</span>, University of Ulm,  gives an invited colloquium <i>Biologically inspired visual-auditory processing – from
+    learn.</blockquote></td></tr><tr valign="top"><td>2022-04-28</td><td>Simon Haller-Seeber contributes a talk <i>STAIR Learning Lab</i> at <a href="http://rie.science">Robotics in Education 2022</a>, online.</td></tr><tr valign="top"><td>2022-02-19</td><td>Justus Piater gives an invited talk <i>Learning as a Creative Process in Humans and
-    brain-like computation to neuromorphic algorithms</i> at <a href="https://www.uibk.ac.at/informatik/forschung/lunchtime-seminar/index.html.en">IFI Lunchtime Seminar</a>. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem11', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem11Abstract" style="display:none">A fundamental task of sensory processing is to detect
+    Machines</i> at <a href="https://www.philosophie.uni-konstanz.de/ag-mueller/online-workshop-on-agency-life-and-creativity/">Workshop
-    and integrate feature items to group them into perceptual units
+    on “Agency, Life, and Creativity”</a>, online. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem5', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem5Abstract" style="display:none">Humans constantly engage in creative activity such as
-    segregating them from other objects and the background. A
+    finding explanations, improving technology, or solving novel
-    framework is discussed which explains how perceptual grouping at
+    problems in everyday life. Even imitating other humans, e.g. while
-    early as well as higher-level cognitive stages may be implemented
+    learning a new skill, is creative in that the imitator has to
-    in cortex. Different grouping mechanisms are implemented which are
+    infer the conceptual structure underlying the observed
-    attuned to basic features and feature combinations and mainly
+    manifestation and to regenerate the latter from the former. In
-    evaluated along the forward sweep of stimulus processing. However,
+    contrast, current AI-enabled agents are unable to generate and
-    due to limitations of local feature detection mechanisms and
+    draw upon such conceptual structures. Their learning is mostly
-    inherent ambiguities, top-down feedback is required to deliver
+    driven by statistics, and any ability to produce novelty is
-    contextual information helping to disambiguate initial
+    largely limited to trial and error. I will try to characterize
-    measurements. Feedback of contextual information is demonstrated
+    this fundamental difference between learning in human and
-    to improve object recognition performance, stabilize learning of
+    artificial agents, and will discuss possible technological
-    object categories, and integrate multi-sensory representations.
+    approaches that might help close this gap.</blockquote></td></tr><tr valign="top"><td>2021-11-24</td><td>Justus Piater is a panelist at a public discussion on <a href="https://www.linkedin.com/posts/elsa-innsbruck_letzte-woche-fand-zum-elsa-day-am-24112021-activity-6871397038093869057-16-7">Privacy
+    in the Digital Age</a>, online. (Organized by ELSA Innsbruck)</td></tr><tr valign="top"><td>2021-10-22</td><td>Justus Piater gives an invited talk <i>Robotik und KI in der Medizin – nur
-    The canonical principles of neural computation define a set of
+    Bedarfserweckung?</i> at <a href="https://medical-update-hall.com/veranstaltungen/">Medical
-    core operations to implement above-mentioned mechanisms of
+    Update Hall 2021 – Fit für das neue Jahrzehnt</a>, UMIT, Hall in Tirol.</td></tr><tr valign="top"><td>2021-08-26</td><td>Justus Piater contributes a talk <i>Digital Science at the University of Innsbruck</i> at Aurora Research Conference Digital Society &amp; Global Citizenship, Vrije Universiteit Amsterdam. <span class="actions"><a href="javascript:void(0)" onclick="showHide('newsitem8', 'Abstract')">[Abstract]</a></span><blockquote id="newsitem8Abstract" style="display:none">Growing data processing capacities and progress in
-    perceptual and cognitive inference. These operations can be
+    analytical methods and artificial intelligence are motivating
-    mapped, in a simplified form, onto neuromorphic platforms to
+    entire branches of science to raise new questions and to develop
-    emulate brain-like computation. It is demonstrated that an
+    new approaches. This digital transformation requires progressive,
-    architecture composed of canonical circuit mechanisms can be
+    interdisciplinary synergies between the computational sciences and
-    mapped onto neuromorphic chip technology facilitating low-energy
+    all other scientific disciplines.  The Digital Science Center
-    non-von Neumann computation.</blockquote></td></tr></table></HTML>
+    (DiSC) at the University of Innsbruck meets this challenge by
+    integrating and promoting digitalisation of scientific research.
+    In this talk I will present some examples of interdisciplinary
+    research by DiSC staff researchers who are experts in specific
+    aspects of digital science.</blockquote></td></tr><tr valign="top"><td>2021-08-24</td><td>Justus Piater teaches a tutorial <i>Picking and Placing</i> at <a href="https://www.uibk.ac.at/informatik/forschung/gmar-robotics-school-2021.html">2021 GMAR Summer School</a>, Villa Blanka, Innsbruck.</td></tr></table></HTML>
 [[news|Older News]]

User Tools

Site Tools

Differences

Page Tools