Ambient Intelligence

Ambient intelligence aims at increasing the quality of life by introducing the technology into our everyday environment in a way that places minimal burden on the user. Due to the rapid aging of the population, care for the elderly and telemedicine are among the area’s main goals, but applications for the general population, such as smart buildings, are also important. A prerequisite for ambient intelligence is understanding the user’s situation and needs, which we address by analyzing human behavior using various sensors. We work with many sensor types, but have particularly extensive experience with:

  • Smartphone sensors
  • Dedicated inertial sensors
  • Real-time location systems
  • Infrared motion capture

We work on a wide range of ambient intelligence tasks, for example:

  • Activity recognition
  • Human energy expenditure estimation
  • Fall detection
  • Detection of unusual behavior caused by health or security issues
  • Detection of unusual environmental events
  • Recognition of diseases
  • Smart home applications focused on energy efficienty, security and ease of use

Contact: M. Luštrek (

Research Projects

  • Fit4Work
    The project is developing tools to monitor physical and mental stress of older workers, and to help them exercise physically and relieve stress to improve health, wellbeing and the ability to work.
  • Commodity12
    The project aims at improving the daily management of diabetes and the prevention/management of its cardiovascular co-morbidities.
  • e-Gibalec
    The project is developing a mobile application to encourege physical activity of schoolchildren using gamification techniques.
  • E-doorman
    The aim of the project is to develop the e-doorman: an intelligent system using door with electro-mechanic lock, tablet PC, and an array of sensors that offers services similar to a human doorman, improves security and increases user comfort.

Past Projects

Doctoral Research Projects

Multi-classifier adaptive training algorithm for learning from unlabelled data

Božidara Cvetković , supervisor: Mitja Luštrek


Nowadays, the applications and systems generate large amounts of unlabelled data which by itself carry no information. For example, wearable and environmental sensors continuously sense and measure the respective data, speech while talking to the mobile assistants (Siri, Google assistant, Alexa, etc.) is sensed, large amount of new content on websites is generated daily, etc. Labelling such large amounts of data is very labour intensive and expensive. Even though the idea and need for automatic labelling appeared decades ago (Scudder 1965, Vapnik and Chervonenkis 1974), this area is still not mature enough to be generally applied.

Semi-supervised learning is a technique that aims at utilising a small set of labelled data to automatically label the unlabelled data. The topic is very popular due to the mentioned fact that in all domains there are cases where large amount of unlabelled data is generated and the need for labelling it exists. There are several approaches to tackle the task but they are either limited to offline labelling or the selection process is implemented with empirical risk minimisation approach, which means that only the most confident predictions are included in retraining and consequently do not contribute effectively to the solution.

The thesis will explore the possibility to employ machine-learning into the selection process and enable inclusion of the predictions which might be discarded using the heuristics approach to empirical risk minimisation. This approach will contribute to inclusion on more diverse data and therefore more effective adaptation of the learner.

Journal papers:

CVETKOVIĆ, Božidara, JANKO Vito, ROMERO E. Alfonso, KAFALI Ozfur, STATHIS Kostas, LUŠTREK, Mitja. Activity Recognition for Diabetic Patients Using a Smartphone. Journal of Medical Systems, 2016, vol. 40, no. 12, str. 256-1-256-8. link

CVETKOVIĆ, Božidara, KALUŽA, Boštjan, LUŠTREK, Mitja, GAMS, Matjaž. Adapting Activity Recognition to a Person with Multi-Classifier Adaptive Training. Journal of Ambient Intelligence and Smart Environments, vol. 7, no. 2, 2015, pp. 171-185. link