We are happy to announce the 2nd HdM Deeplearning Day at Friday, January 12th, 2018, room I003, Nobelstr. 8
Deep Learning is currently one of the hottest topics in computer science. The performance of existing applications such as object recognition, speech recognition or automatic translation have been revolutionized by deep neural networks. Moreover, this type of algorithms open doors for a wide range of new intelligent applications.
The presentations of the 2nd Deeplearning cover current research directions, applications of deep learning in autonomous driving and architecture of complex and high-performing AI-systems.
Schedule:
13.30h: Welcome (Prof. Dr. Johannes Maucher, HdM)
13.45h: Understanding the world by learning how to model it (Johannes Theodoridis, M.Sc., HdM) - Artificial Intelligence is on everyone’s lips. But how “intelligent” are our AI systems really? What is resilient about the current hype, where do we stand and what can we expect? A deep dive into the latest developments in AI and machine learning.
14.45h: Machine Learning for Autonomous Driving (Michael Herman, Dr. Bastian Bischoff, Robert Bosch GmbH, Differentiating AI - Environmental Understanding - Decision Making) - For bringing autonomous vehicles on public roads, many questions need to be answered. For example, what are essential features in the high-dimensional observations from varying sensors, what are other traffic participants going to do, how are they going to react to the behavior of the autonomous car, how should an autonomous system behave, or how to find optimal strategies according to these criteria, while guaranteeing safety requirements. Machine Learning approaches can be used to answer these questions to some extent. While Deep Neural Networks have outperformed traditional methods in various applications, they have also shown to be vulnerable against adversarial perturbations on problems with high-dimensional input spaces. Since this can prevent its usage in safety- and security-critical applications such as autonomous driving, it is necessary to both understand the limits of these models and to increase their robustness.
15.45h: Applications of Deep Learning in Autonomous Driving (Eike Rehder, Daimler AG, R&D – Environment Perception) - Deeplearning is considered one of the enabler technologies for automated driving. Due to the cognitive versatility of neural networks, they can be applied in nearly every component of automated driving. In this talk, we will demonstrate examples of perception, situation interpretation and decision making for automated vehicles.
16.45h: Cognitive Systems (Dr. Thilo Maurer, IBM Power Acceleration and Design Center) - IBM classifies applications of artificial intelligence, expert algorithms and robotics as cognitive systems. This class differs significantly from traditional IT applications in that it also uses unstructured data such as images, text, speech and sensor data. Economical processing of such thin data has only become possible in recent years due to the progressive reduction in the price of computing units. In this context, methods oriented to the functioning of the human brain are increasingly being used. The lecture will also discuss the IBM Watson Software, current machine learning hardware, and the SyNAPSE research project, and how these approaches fall under artificial intelligence.