Dependable Deep Learning

Dependable Deep Learning

Deep Learning systems perform surprisingly well on a broad set of tasks from various domains, including computer vision and natural language processing. Recently, however, theoretical studies, as well as, for example, prominent cases of accidents involving self driving vehicles, which received broad coverage by the media, demonstrated that Deep Learning based models sometimes fail, seemingly without any reason.

Read More
optiPlan – Robot Trajectory Optimization for Relaxed Effective Tasks

optiPlan – Robot Trajectory Optimization for Relaxed Effective Tasks

Many industrial effective tasks, e.g., welding, cutting, allow a certain freedom during their execution, e.g., the robot’s tool might have a certain deviation during welding. This freedom can be used as an extra degree of freedom for robot trajectory optimization. In this project, we formalize these tasks and optimize robot trajectory.

Read More
checkAll

checkAll

In modern software and systems engineering formal verification gets more and more important to ensure safeness of safety critical systems. Especially, if the systems get in contact with human beings. Therefore, we research for new techniques in the domain of formal verification for software intensive systems.

Read More
Fine-Grained Visual Recognition of Retail Products

Fine-Grained Visual Recognition of Retail Products

Open set recognition at scale is an issue for computer-vision systems, especially in the retail industry in which scalable AI-related technology solutions at the physical store level have not been deployed to the best of our knowledge.

Read More
VECS

VECS

VECS is an Eclipse-based specification framework for the analysis of software-intensive systems. It includes an implementation of the SAML language and model transformations to verification engines. Currently, connectors to high-level Software Engineering tools (Rhapsody) are being developed.

Read More
TouchAuth

TouchAuth

Smart mobile devices are gaining widespread acceptance as displays sizes, computation power and mobile data transfer bandwidth increase. However, this also leads to new security issues. Here, we develop new bio-metric authentification methods. The core idea is to use soft keyboards as a biometric feature for authentification.

Read More
API Specific Automatic Program Repair (ASAP-Repair)

API Specific Automatic Program Repair (ASAP-Repair)

API Specific Automatic Program Repair or how can we find and fix API Misuses automatically? Nowadays, programmers re-use much code from existing code libraries by means of Application Programming Interfaces (APIs). Due to missing or outdated documentation as well as misunderstandings on how to correctly use a particular API, programmers may falsely apply that API.

Read More
Opti-Spec

Opti-Spec

In early stages of software engineering projects, several fundamental decisions must be made. At the same time, there exists a number of potentially antagonistic goals and/or requirements. If these decisions are not simply yes/no decisions, but are numerical in nature, then it is necessary to find an optimal configuration in the design space.

Read More
SafeMotion

SafeMotion

Autonomous mobile robots gain more and more importance. In the nearest future, they will be a part of everyday life. Therefore, it is critical to make them as reliable and safe as possible even in an environment cluttered with unexpectedly moving obstacles.

Read More
Elastic Coverage Path Planning

Elastic Coverage Path Planning

Industrial surface processing by robots - applications such as cleaning, material removal or painting - are characterized by low automation level and consequently remain manually intensive. For automated robot control in such applications we need a solution of the Coverage Path Planning (CPP) problem, which could be performed either off-line or on-line.

Read More