I am an associate professor and the Computer Communications Lab head at the Faculty of Computer and Information Science (FRI), University of Ljubljana, Slovenia. My research deals with mobile computing. Lately I have been focused on mobile sensing and machine learning, approximate mobile computing, and the security in the IoT domain.
Contact:
- email: Veljko.Pejovic [at] fri.uni-lj.si
- post: Room 3.15 Faculty of Computing and Information Science University of Ljubljana, Vecna pot 113, Ljubljana, Slovenia
Researcher positions (internship and PhD) in Approximate Mobile Computing!
I am looking for researchers passionate about making mobile devices more resource efficient to join us at the University of Ljubljana. More info!
News
- Our work won the Outstanding research achievement award for 2024 by the Faculty of Computer and Information Science!
- I am on the Organisation Committee of ACM UbiComp 2025 on the Programme Committee for ACM UMAP and EuroMLSys, and I serve as an Associate editor for ACM IMWUT and ACM JCSS. Please consider submitting your work to these venues.
- June 2024: Andraž's massive effort in collecting multimodal sensor data for IoT-based authentication research has resulted in a large publicly available dataset. Please get in touch, if interested in using this dataset.
- Jan 2024: What is it like working with us? Check out these cool Facebook reels to find out! video 1 and video 2
Recent publications
- On March 31st in Rotterdam you can hear all about our work on a split learning framework at EuroMLSys, as well as our on-device LLM for semantic inference on Android phones at EdgeSys.
- Approximate mobile computing for UAV-based precision agriculture is demonstrated in our latest IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing paper.
- CoLExT - our testbed for federated learning research in collaboration with KAUST and was presented at ACM/IEEE SEC in December 2024. Further experiments on the testbed are described in Boris' ACM Transactions on Modeling and Performance Evaluation of Computing Systems paper.
- Alina's recent paper on adaptation algorithms completes her work on compressive sensing-inspired deep learning for ubiquitous computing. Previously she has developed a novel adaptive processing pipeline and has thoroughly surveyed the domain.
- Mobiprox - a mobile computing compiler suite that enables dynamically adaptable approximate computing is now published by IEEE Internet of Things Journal. This is a joint work with University of Illinois Urbana-Champaign.
- In our latest ACM GetMobile piece we present a vision of an accuracy-tuneable accelerator for mobile deep learning.