Research Projects

Cognitive perceptual models in machine learning of human activities

This project will investigate advances in cognitive theories to develop activity recognition and behaviour inference models that can cope with changing concepts and allow dynamic reconfiguration  of sensor networks at run time.

Seamless integration of wearable computers & public displays for mobile collaboration in hospital

 This project targets the development and evaluation of interaction technology for wearable computers to support mobile collaboration in hospital setting. Specific focus will be paid on mixed-reality session establishment, multimodal interaction, and context-sensitive user interface adaptation based on the situation. A wearable system prototype that provides activity support both at an individual level and in collaborative settings will be designed and evaluated together with hospital partners.

Personal fabrication: supporting new devices and assisted living

This research aims to understand the problems and complexities which for now prohibit a more widespread adoption of personal fabrication, especially in the area of healthcare and assisted living. This knowledge will inform the design of new systems and interfaces, empowering a broader mass to utilize and benefit from personal fabrication.

Out-patient Monitoring with Wearable Sensors

The project aims at developing prototypes and algorithms for the unobtrusive monitoring of stroke survivors in out-patient therapy. Computer aided rehabilitation has so far mainly focussed on augmenting supervised and scripted motor trainings with IuK technology. Yet for comprehensive and fast motor recovery self-motivated in-vivo training, i.e. patients’ efforts to use their impaired limbs as much as they can in everyday life, is crucial.

Human Circadian Phase Estimation using Unobtrusive Wearable Sensor Data

This work focuses on the estimation of circadian phase using heart rate variability (HRV), both in the time and the frequency domain. Given the HRV signal’s proneness to masking effects, careful consideration will be given to the demasking of this signal and the selection of robust modeling techniques.

Early detection of worsening heart failure using continuous non-invasive measurements

The aim of this project is to develop robust algorithms for non-invasive thoracic impedance measurements and to develop both uni- and multi-parametric algorithms to provide accurate and early alarms of worsening HF.

Semi-supervised Machine Learning for Countering Concept Drift in Security Applications

Our project is initially motivated by an access control system making use of Machine Learning to increase security and help with the potential shortcomings of rule based access control system such as missing or conflicting rules.

Embedded eyes: integrate eye-sights into objects

This project will investigate new computer vision and image processing techniques for eye tracking and gaze estimation, as well as develop novel gaze-based interface for spontaneous and casual human-computer interaction.

Long-term activity monitoring in AAL using a large-area sensitive floor

This research project aims to build such a system using SensFloor [2][3], a large-area capacitive underlay, as the main pervasive ambient sensor. SensFloor can provide information about the position and state of a person (walking speed, direction, detection of falls) .

Privacy-respecting programming

Our approach creates a data ownership map across the framework for each user in order to implement the tracking and deletion of personal data. To do so, we represent the code of a program as a set of functions and check for the existence of the corresponding inverse function -1. If -1 does not exist, the output can be stored without compromising the input, we call the function privacy respecting.