Marko Harasic
Marko Harasic has a total of 17 co-authors. Here are the top co-authors by number of shared publications:
- Adrian Paschke
Fraunhofer Institute for Open Communication Systems
3 shared publications - Denny Mattern
Data Analytics Center, Fraunhofer Institute for Open Communication Systems, Berlin, Germany
2 shared publications Arun Prakash
Fraunhofer FOKUS, Berlin, Germany
1 shared publication- Babak Mafakheri
Safran Passenger Innovations, Weling, Germany
1 shared publication - Felix-Sebastian Keese
Data Analytics Center, Fraunhofer Institute for Open Communication Systems, Berlin, Germany
1 shared publication - Filippo Mazza
Applied Network Technology Lab, Huawei Technologies Duesseldorf GmbH, Riesstrasse 25, 80992 Munich, Germany
1 shared publication
Latest Publications
- 2 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 192-199
Neural Network-Powered Finger-Drawn Biometric Authentication
This paper investigates neural network-based bio-metric authentication using finger-drawn digits on touchscreen devices. We evaluated CNN and auto encoder architectures for user authentication through simple digit patterns (0–9) traced with finger input. Twenty participants contributed 2,000 finger-drawn digits each on personal touchscreen devices. We compared two CNN architectures: a modified Inception-V1 network and a lightweight shallow CNN for mobile environments. Additionally, we examined Convolutional and Fully Connected autoencoders for anomaly detection. Both CNN architectures achieved ~89% authentication accuracy, with the shallow CNN requiring fewer parameters. Autoencoder approaches achieved ~75% accuracy. The results demonstrate that finger-drawn symbol authentication provides a viable, secure, and user-friendly biometric solution for touchscreen devices. This approach can be integrated with existing pattern-based authentication methods to create multi-layered security systems for mobile applications.
- 2 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 192-199
Neural Network-Powered Finger-Drawn Biometric Authentication
This paper investigates neural network-based bio-metric authentication using finger-drawn digits on touchscreen devices. We evaluated CNN and auto encoder architectures for user authentication through simple digit patterns (0–9) traced with finger input. Twenty participants contributed 2,000 finger-drawn digits each on personal touchscreen devices. We compared two CNN architectures: a modified Inception-V1 network and a lightweight shallow CNN for mobile environments. Additionally, we examined Convolutional and Fully Connected autoencoders for anomaly detection. Both CNN architectures achieved ~89% authentication accuracy, with the shallow CNN requiring fewer parameters. Autoencoder approaches achieved ~75% accuracy. The results demonstrate that finger-drawn symbol authentication provides a viable, secure, and user-friendly biometric solution for touchscreen devices. This approach can be integrated with existing pattern-based authentication methods to create multi-layered security systems for mobile applications.
- 22 November 2025
- Published by Springer Nature
- p. 289-300
Terminologie-Editor
- 20 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Aerospace and Electronic Systems Magazine
- Vol. 38 (12) , 14-27
AI-Powered Edge-Cloud Continuum for In-Flight Entertainment and Connectivity
The aviation industry is moving towards a greener, more sustainable, integrated, and digital ecosystem, with Artificial Intelligence (AI) and Machine Learning (ML) showing potential key roles in the transformation process. Travelers, who are now increasingly used to ubiquitous data access, are restarting their air travels with higher expectations and demands for connectivity services. This trend highlights the importance of the In-Flight Entertainment and Connectivity (IFEC) system, which should adapt to such changes and be designed to focus on network security and privacy. In this article, we provide experimental demonstration of an AIbased edge-computing platform developed within the Cloudenabled Aircraft Network and ARtificial Intelligence-based data Analysis (CANARIA) project, which targets to deliver proofof- concept of an in-flight edge network. The CANARIA edgecomputing platform offers a set of AI-based and containerized applications to not only improve the in-flight experience for cabin crew and passengers, but also to underpin the cabin digital transformation while increasing the safety and security of the connectivity system.
- 20 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Aerospace and Electronic Systems Magazine
- Vol. 38 (12) , 14-27
AI-Powered Edge-Cloud Continuum for In-Flight Entertainment and Connectivity
The aviation industry is moving towards a greener, more sustainable, integrated, and digital ecosystem, with Artificial Intelligence (AI) and Machine Learning (ML) showing potential key roles in the transformation process. Travelers, who are now increasingly used to ubiquitous data access, are restarting their air travels with higher expectations and demands for connectivity services. This trend highlights the importance of the In-Flight Entertainment and Connectivity (IFEC) system, which should adapt to such changes and be designed to focus on network security and privacy. In this article, we provide experimental demonstration of an AIbased edge-computing platform developed within the Cloudenabled Aircraft Network and ARtificial Intelligence-based data Analysis (CANARIA) project, which targets to deliver proofof- concept of an in-flight edge network. The CANARIA edgecomputing platform offers a set of AI-based and containerized applications to not only improve the in-flight experience for cabin crew and passengers, but also to underpin the cabin digital transformation while increasing the safety and security of the connectivity system.
- Open Access
- 25 August 2023
- Published by Springer Nature in International Journal of Data Science and Analytics
- Vol. 17 (4) , 337-357
Recent advances and future challenges in federated recommender systems
Recommender systems are an integral part of modern-day user experience. They understand their preferences and support them in discovering meaningful content by creating personalized recommendations. With governmental regulations and growing users’ privacy awareness, capturing the required data is a challenging task today. Federated learning is a novel approach for distributed machine learning, which keeps users’ privacy in mind. In federated learning, the participating peers train a global model together, but personal data never leave the device or silo. Recently, the combination of recommender systems and federated learning gained a growing interest in the research community. A new recommender type named federated recommender system was created. This survey presents a comprehensive overview of current research in that field, including federated algorithms, architectural designs, and privacy mechanisms in the federated setting. Furthermore, it points out recent challenges and interesting future directions for further research.
- 25 August 2023
- Published by Springer Nature in International Journal of Data Science and Analytics
- Vol. 17 (4) , 337-357
Open AccessRecent advances and future challenges in federated recommender systems
Recommender systems are an integral part of modern-day user experience. They understand their preferences and support them in discovering meaningful content by creating personalized recommendations. With governmental regulations and growing users’ privacy awareness, capturing the required data is a challenging task today. Federated learning is a novel approach for distributed machine learning, which keeps users’ privacy in mind. In federated learning, the participating peers train a global model together, but personal data never leave the device or silo. Recently, the combination of recommender systems and federated learning gained a growing interest in the research community. A new recommender type named federated recommender system was created. This survey presents a comprehensive overview of current research in that field, including federated algorithms, architectural designs, and privacy mechanisms in the federated setting. Furthermore, it points out recent challenges and interesting future directions for further research.
- Open Access
- 17 August 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (16) , 8239
A Review on AI for Smart Manufacturing: Deep Learning Challenges and Solutions
Artificial intelligence (AI) has been successfully applied in industry for decades, ranging from the emergence of expert systems in the 1960s to the wide popularity of deep learning today. In particular, inexpensive computing and storage infrastructures have moved data-driven AI methods into the spotlight to aid the increasingly complex manufacturing processes. Despite the recent proverbial hype, however, there still exist non-negligible challenges when applying AI to smart manufacturing applications. As far as we know, there exists no work in the literature that summarizes and reviews the related works for these challenges. This paper provides an executive summary on AI techniques for non-experts with a focus on deep learning and then discusses the open issues around data quality, data secrecy, and AI safety that are significant for fully automated industrial AI systems. For each challenge, we present the state-of-the-art techniques that provide promising building blocks for holistic industrial AI solutions and the respective industrial use cases from several domains in order to better provide a concrete view of these techniques. All the examples we reviewed were published in the recent ten years. We hope this paper can provide the readers with a reference for further studying the related problems.
- 17 August 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (16) , 8239
Open AccessA Review on AI for Smart Manufacturing: Deep Learning Challenges and Solutions
Artificial intelligence (AI) has been successfully applied in industry for decades, ranging from the emergence of expert systems in the 1960s to the wide popularity of deep learning today. In particular, inexpensive computing and storage infrastructures have moved data-driven AI methods into the spotlight to aid the increasingly complex manufacturing processes. Despite the recent proverbial hype, however, there still exist non-negligible challenges when applying AI to smart manufacturing applications. As far as we know, there exists no work in the literature that summarizes and reviews the related works for these challenges. This paper provides an executive summary on AI techniques for non-experts with a focus on deep learning and then discusses the open issues around data quality, data secrecy, and AI safety that are significant for fully automated industrial AI systems. For each challenge, we present the state-of-the-art techniques that provide promising building blocks for holistic industrial AI solutions and the respective industrial use cases from several domains in order to better provide a concrete view of these techniques. All the examples we reviewed were published in the recent ten years. We hope this paper can provide the readers with a reference for further studying the related problems.