Machine Vision for HCI

Researchers, engineers, graduate and undergraduate students interested in developing machine vision systems aimed at man‐machine interaction, all this without the need to code everything from scratch.

Machine vision finds applications in many domains from medicine, archaeology, oceanography to robotics. However, due to the relative low cost of image capturing devices and the plethora of information contained in the visual domain, machine vision is migrating to other fields too, making it in particular suitable for developing human‐computer interaction scenarios. Machine vision can enable detection, tracking, identification and interpretation of actions performed by humans, making it at the same time an intuitive and natural way of accepting commands from users. Developing image processing algorithms and then coding them using complex software libraries like OpenCV can be a daunting task for many researchers, especially if results are desired quickly; even though software like Matlab provides the ability to rapidly prototype image processing systems, their real‐time performance is poor and the flexibility to integrate these into interaction projects is not high. Participants will also have the opportunity to get an overview of algorithms from image processing, which can be used to develop visual human‐computer interaction scenarios. Following this, they will be introduced to RoboRealm, a powerful vision software application that allows for rapid development of machine vision systems with little to no programming required.

Tutorial: A Practical Approach to Machine Vision for Robot Control – Accepted at the 18th Mediterranean Conference on Control and Automation (MED 2010)

IURS Summer schools on Smart Systems (including topics in Computer Vision) in 2007, 2008, 2009.

Teaching courses of Human‐Computer Interaction and Computer Vision at TU Kaiserslautern and Jacobs University Bremen.

Other summer schools and tutorials.