In the CVMLS group, we specialize in data science methodologies, including artificial intelligence, deep learning, computer vision, and image processing. Besides researching the theoretical aspects of these algorithms, we are intrigued by the inclusion of new and/or customized hardware components, such as state-of-the-art sensors.
This multi-level combination is particularly promising for bridging the gap between different scientific disciplines and identifying novel solutions to technically challenging and socially important questions. Our ultimate research goal is to enhance understanding while developing innovative solutions with real-world applicability.
Computers are arguably the most versatile man-made technology of our time. These universal machines are deeply integrated into our society, and their transversal nature impacts all academic and non-academic disciplines alike. In particular, the ability of computers to process and analyze visual data (computer vision) and to automatically identify patterns within data (machine learning) enables them to perform complex intellectual tasks (colloquially referred to as artificial intelligence).
Despite the recent successes of computer vision (CV) and machine learning (ML), applying these systems to real-world problems remains a significant challenge. In our group, we address these challenges by investigating the theoretical limitations of modern CV and ML algorithms to derive abstract and generalizable solutions. Ideally, these solutions contribute to interdisciplinary applications, improving automated and high-throughput data analyses while facilitating the transfer of these technologies to industry and society.
In addition to the algorithmic focus of our group, we are also interested in state-of-the-art hardware systems (S), such as innovative imaging sensors and advancements in 3D printing. Indeed, only by considering the entire pipeline—from data acquisition and processing to data output and actuation—can we pave the way for comprehensive, data-driven, and sustainable technologies. This systematic approach defines our research as Computer Vision & Machine Learning Systems (CVMLS).
Given the universal character of computers in general and modern deep learning and computer vision algorithms in particular we are collaborating with a vareity of interdisciplinary partners.
