Research Team

I am fortunate to work with a fantastic group of people at University of Ljubljana and beyond:

• Dr. Octavian Machidon, postdoc, area: approximate mobile computing
• Dr. Alina Machidon, postdoc, area: compressive sensing for mobile computing
• Andraž Krašovec, PhD student (collaborative doctoral programme with JRC) thesis topic: Sensor-based authentication
• Boris Radovič, PhD student (collaborative doctoral programme with KAUST) thesis topic: Context-aware federated learning
• Christoph Anderson, PhD student (coadvisor), Kassel University, Germany
• Aneta Kartali, PhD student (advised by Dr. Octavian Machidon)
• Timotej Ciglarič, MS, thesis topic: Approximate computing for resource-efficient on-device neural network training
• Ilina Kirovska, MS, thesis topic: Analysis of the Capabilities of Large Language Models for Mental Health Inference from Mobile Sensor Data
• Martin Korelič, MS, thesis topic: Parameter-Efficient Tuning of Large Language Models on Mobile Devices

Research Interests

• Resource-efficient mobile and IoT computing
• Approximate mobile deep learning
• Federated learning with on-device training implementation
• Sustainable computing
• Ubiquitious computing sensing and machine learning for various applications (cybersecurty above all)

Recent Key Funding/Projects

approXimation for adaptable diStributed artificial intelligence (ARIS, PI, 2025-2027, 150,000 EUR)

CODA - "Context-Aware On-Device Approximate Computing" (ARRS, PI, 2021-2025, 300,000 EUR)

AgriAdapt - Adaptable Deep Learning for Computer Vision-Based Agriculture (H2020, co-PI, 2023, 80,000 EUR)

CARMA - "Bringing Resource Efficiency to Smartphones with Approximate Computing" (ARRS, PI, 2020-2022, 200,000 EUR)

Behaviour Biometrics for Authentication in IoT Environments (Joint PhD Programme with European Commission, 2020-2024)

RICERCANDO - Rapid Interpretation and Cross-Experiment Root-Cause Analysis in Network Data with Orange (H2020, co-PI, 2016-2018 150,000 EUR)

AKOS - Feasability study: Crowdsourced measurements mining for Net neutrality violation detection (Industrial, PI, 2018, 20,000 EUR)

Mobile computing for efficient information dissemination in emergency situations (US-Slovenia bilateral project, PI, 2020-2021, 3,000 EUR)

Publications

All publications are listed here.

Lab alumni

• Brin Colnar, MS, thesis title: Reinforcement learning for efficient UAV-based computer vision, defended October 2024.
• Janez Božič, MS, thesis title: Design and implementation of a heterogeneous federated learning testbed, defended November 2024.
• Boris Radovič, MS, thesis title: Client Clustering for Improved Federated Learning on Heterogeneous Data", defended September 2023.
• Matevž Fabjančič, MS, thesis title: "Compiler-level approximate mobile computing", defended October 2021.
• Timotej Knez, MS, thesis title: "Context-dependent adjustment of computation complexity in mobile sensing", defended September 2021.
• Andraž Krašovec, MS, thesis title: "Continuous Authentication in IoT Environments", defended September 2020.
• Jasna Urbančič, MS, thesis title: "Transportation Mode Detection Based on Mobile Sensor Data", co-advised with Dr Dunja Mladenič, defended December 2018.
• Tilen Matkovič, MS, thesis title: "Inferring Cognitive Load using Wireless Signals", defended November 2018.
• Gašper Urh, MS, thesis title: "Mobile Sensing for Task Engagement Inference", defended October 2016.
• Nejc Bizjak, MS, thesis title: "Context Prediction-Based Prefetching in Software-Defined Wireless Networks", defended October 2016.
• Anže Kocjančič, BS, thesis topic: Federated learning for interruptiblity inference in mobile computing, defended September 2024.
• Blaž Kristan, BS, thesis topic: approximate mobile computing for efficient image processing, defended September 2022.
• Domen Vilar, BS, thesis topic: cross-country skiing safety app, defended September 2022.
• Jani Asprov, BS, thesis title: "Harnessing human activity sensing for resource-efficient mobile video adaptation", defended October 2021.
• Anze Kristan, BS, thesis title: "Examining the potential of autoencoders for automatic feature extraction in wireless cognitive load inference", defended September 2021.
• Janez Bozic, BS, thesis title: "Approximate mobile computing with loop perforation", defended September 2021.
• Emir Hasanbegović, BS, thesis title: "The Role of Mobile Device Context in Perceiving News App View Layouts", defended February 2021.
• Andrej Martinovič, BS, thesis title: "The Analysis of the Role of the Context on Mobile Ad Perception", defended October 2020.
• Aleksandar Cuculoski, BS, thesis title: "Energy-Efficient Intelligent Mobile Notification Management", defended August 2020.
• Amadej Janković, BS, thesis title: "Personalised push notifications for motivating behaviour change through a mobile application", defended September 2020.
• Tine Fajfar, BS, thesis title: "Resource consumption analysis of approximate video decoding for mobile devices", defended April 2020.
• Aljaž Verlič, BS, thesis title: "Anomaly Detection in Mobile Broadband Measurements", defended December 2019.
• Timotej Knez, BS, thesis title: "Analysis of Physiological Signal Elicited by Mobile Gaming", defended September 2019.
• Martin Frlin, BS, thesis title: "Developing Software Tools for In Situ Cognitive Load Estimation", defended March 2019.
• Rosanda Potrebuješ, BS, thesis title: "Mobile Application for Context Inference at Entertainment Events", defended March 2019.
• Ziga Kljun, BS, thesis title: "Mobile computing support for warehouses", defended August 2018.
• Marko Ceferin, BS, thesis title: "Specification and Implementation of a Light-Weight Internet Content Delivery Platform", defended October 2017.
• Dongho Kim, visiting student from KNU, South Korea, area: approximate mobile computing, summer 2019
• Ivan Majhen, research assistant, area: mobile computing, power and network measurement analysis, 2017-2020
• Luna Živkovic, visiting MS from FTN, Serbia, area: ESM framework for mobile approximation acceptance assessment, summer 2022
• Jhihoon Joo, visiting PhD student from KNU, South Korea, area: vehicle-to-vehicle wireless channel analysis and prediction, spring 2018