Tutorial (September 10, Morning):

Tutorial: Automated Machine Learning: Introduction to Hyperparameter Optimization and Neural Architecture Search

Matthias Feurer und Thomas Elsken (Univ. of Freiburg, Germany)

The success of machine learning crucially relies on human machine learning experts, who construct appropriate features and workflows, and select appropriate machine learning paradigms, algorithms, neural architectures, and their hyperparameters. Automated Machine Learning (AutoML) is an emerging research area that targets the progressive automation of machine learning, which uses machine learning and optimization to develop off-the-shelf machine learning methods. It targets both ML researchers and non-ML experts, easing and enabling the use of machine learning algorithms. AutoML covers a broad range of subfields, including hyperparameter optimization, neural architecture search and meta-learning. This tutorial will cover the methods underlying the current state of the art in this fast-paced field and also contain hands-on exercises using state-of-the-art AutoML tools.

Outline (tentative):

  • Introduction & Motivation
  • Automated Machine Learning via Hyperparameter optimization
  • Neural Architecture Search
  • Meta-Learning & Learning to Learn
  • Hands-on


Tutorial (September 10, Afternoon):

Tutorial: Vision for robotics

Tim Patten (TU Vienna, Austria), Cesar Cadena (ETH Zurich, Switzerland)

Robot perception is the act of interpreting sensor data to generate an awareness of the surrounding environment and robot vision specifically interprets the data from onboard cameras. In recent years, robot vision has experienced significant leaps forward due to the availability of inexpensive RGB-D cameras, which has enabled direct perception of the 3D world, and the exploitation of deep learning, which has established state of the art for many semantic tasks. Despite these progressions, there is a frustrating performance gap between computer vision algorithms tested in the lab and those deployed in the wild.

The aim of this tutorial is to cover the concepts, methods, applications and challenges of vision for robotics in order to both expose and bridge the gap between computer and robot vision. The tutorial will be divided into two sessions. The aim of the first session is to give an overview of simple robotic platforms that are suitable for different tasks to encourage computer vision experts to test their work in real-world domains. The second session will cover semantic understanding of the environment with a focus on modern deep learning approaches that are frequently applied in robotics.

Part I: Which robotic platform should I use?

This talk will look in detail at available robot platforms that are suitable for robotic vision experimentation. The focus will be on out-of-the-box platforms that can be simply set up and used directly to deploy vision algorithms. This will include an overview of system prerequisites, software frameworks and a description of common platforms that are usable with minimal hardware knowledge.

Part II: Semantics and deep learning for robotic perception

This talk will look at the typical vision tasks faced by robots and clarify the differences between methods and results when algorithms are applied in the lab as opposed to the real world. This will first cover the challenges of employing standard techniques and then give an overview of frequently used methods in the robotics context by describing what does and does not work. The talk will also present state of the art for vision tasks, such as object detection, as well as robotic related tasks, such as grasp point estimation.



Seminar (September 13, Afternoon):
René Grzeszick (MotionMiners GmbH, Dortmund, Germany)

From Research to Practice – Founding a Start-up after the Doctorate

Focus of the seminar is the thematic framework “From Research to Practice” and the creation of a
start-up company out of a research topic. The seminar follows the example of an EXIST Transfer of
Research in the field of machine learning at the Fraunhofer IML/TU Dortmund University. It describes
the path starting at a PhD topic at a university or research institute until the first sold product. The
seminar addresses PhD students, which are interested in founding their own company, as well as
professors that would like to support a start-up project at their chair.

Note: As the seminar is built around funding and possibilities in Germany, it will be held in German.

Von der Forschung in die Praxis – Gründen nach der Promotion

Schwerpunkt des Seminars ist der Themenrahmen „Von der Forschung in die Praxis“, am Beispiel eines
realisierten EXIST Forschungstransfers im Bereich maschinelles Lernen am Fraunhofer IML/an der TU
Dortmund. Dabei wird er Weg von der Entstehung einer Forschungsidee an der Hochschule oder
Forschungseinrichtung bis hin zum ersten Produktverkauf beschrieben. Das Seminar richtet sich
sowohl an gründungsinteressierte Doktoranden, sowie an Lehrstuhlinhaber, die ein Gründungsprojekt
an ihrem Lehrstuhl begleiten möchten.

Das Seminar beginnt mit Grundlagen der Geschäftsmodellentwicklung, mit besonderem Fokus auf
digitalen Geschäftsmodellen mit Hilfe künstlicher Intelligenz. Dabei werden Methoden wie das
Businessmodell Canvas und einfache Methoden zum Aufbau von Preismodellen für datengetriebene
Dienstleistungen vorgestellt. Es werden Beispiele mehrerer KI Start-Ups aus Deutschland diskutiert.

Anschließend wird das Thema Finanzierung von Ausgründungen beleuchtet. Hierbei wird die
Förderlandschaft in Deutschland vorgestellt, von Bundes- und Landesförderprojekten, von denen
Lehrstühle oder Forschungseinrichtungen, sowie Gründer profitieren können, bis hin zu Möglichkeiten
der (pre-)Seed Finanzierung durch Risikokapitalgeber.

Für die erfolgreiche Durchführung eines Gründungsvorhabens an der Hochschule werden praktische
Erfahrungsbeispiele gegeben, von technischen Entwicklungshürden, den Kontakt mit ersten
potentiellen Kunden, bis hin zu rechtliche Fallstricken. Am Ende eines Gründungsprojektes steht dann
die Ausgründung eines Unternehmens aus der Hochschule bzw. der Forschungseinrichtung an. Die
wichtigsten Punkte auf dem Weg dorthin, sowie die anschließende Gewinnung erster Pilotkunden,
werden kurz diskutiert.

Am Ende des Seminars haben die Teilnehmer einen Überblick darüber gewonnen, wie aus einer
Forschungsidee oder einem Promotionsvorhaben ein Geschäftsmodell und schlussendlich eine
Ausgründung entwickelt werden kann.