@conference{2019_iwssip,
title = {Deep Learning vs. Safety - Practical Approach and Platform Design Perspective},
author = {Milan Z. Bjelica},
url = {http://www.milanbjelica.info/wp-content/uploads/2019/06/program-web.pdf
https://www.youtube.com/watch?v=tJrDJsBiDqg},
year = {2019},
date = {2019-06-06},
booktitle = {Proceedings of 2019 International Conference on Systems, Signals and Image Processing (IWSSIP)},
publisher = {EURASIP},
address = {Osijek, Croatia},
abstract = {Deep Learning is a promising field, allowing an increase in artificial intelligence applications across many fields, ranging from data science, medical, weather, and aerospace to automotive. Applications of computer vision-based deep learning are vastly assisted by modern System-on-Chip architectures, which provide the required parallelism, heterogeneity and interfacing. However, the application of deep learning to safety-critical contexts where human lives might be at stake, such as in self-driving cars, still has many pitfalls. Ongoing academic research tackles transparent AI, in which the correctness of AI is attempted to be reached by design; however, the outcome of this research is still far-fetched. In this talk, we will discuss a practical approach when integrating deep learning vision-based solutions into a safety-critical context, which can be achieved today. We outline an approach which introduces a software/hardware platform design which fosters diversity, with the goal of minimizing risk of critical failures which are induced by AI in decision making.},
howpublished = {M32},
note = {Keynote},
keywords = {ADAS, automotive, deep learning, keynote, market research},
pubstate = {published},
tppubtype = {conference}
}
Deep Learning is a promising field, allowing an increase in artificial intelligence applications across many fields, ranging from data science, medical, weather, and aerospace to automotive. Applications of computer vision-based deep learning are vastly assisted by modern System-on-Chip architectures, which provide the required parallelism, heterogeneity and interfacing. However, the application of deep learning to safety-critical contexts where human lives might be at stake, such as in self-driving cars, still has many pitfalls. Ongoing academic research tackles transparent AI, in which the correctness of AI is attempted to be reached by design; however, the outcome of this research is still far-fetched. In this talk, we will discuss a practical approach when integrating deep learning vision-based solutions into a safety-critical context, which can be achieved today. We outline an approach which introduces a software/hardware platform design which fosters diversity, with the goal of minimizing risk of critical failures which are induced by AI in decision making.
@conference{ceskeynote1,
title = {Smart Devices, Homes, Cars: Is the Modern Consumer Overwhelmed?},
author = {Milan Z. Bjelica},
url = {http://www.icce.org/conference/icce-2017-special-sessions/
https://www.youtube.com/watch?v=KK7CwKT1qNU},
year = {2017},
date = {2017-01-09},
urldate = {2017-01-09},
booktitle = {IEEE International Conference on Consumer Electronics},
abstract = {A potpourri of modern day technologies, which assume connectivity and ubiquitous access, accelerate the pace of our everyday lives. Although useful at the first glance, the galore of choices keep users entangled and hard on new technology adoption. This is particularly true with the adoption of recent devices and applications, including, for example, smart watches, internet of things and home automation technologies. It may be a daunting task for the consumer electronics industry to devise a specific, targeted and clear use cases which would promote the technology. In this talk we would give an outline of current challenges in this regard, via the examples of current products and R&D endeavors within the RT-RK research institute. },
howpublished = {M32},
note = {Keynote},
keywords = {business, internet of things, keynote, market research, smart homes},
pubstate = {published},
tppubtype = {conference}
}
A potpourri of modern day technologies, which assume connectivity and ubiquitous access, accelerate the pace of our everyday lives. Although useful at the first glance, the galore of choices keep users entangled and hard on new technology adoption. This is particularly true with the adoption of recent devices and applications, including, for example, smart watches, internet of things and home automation technologies. It may be a daunting task for the consumer electronics industry to devise a specific, targeted and clear use cases which would promote the technology. In this talk we would give an outline of current challenges in this regard, via the examples of current products and R&D endeavors within the RT-RK research institute.
@conference{indel1,
title = {The Personalization of a Cloud Ecosystem: Adding Dimensions to Situational Awareness},
author = {Gordana Velikić and Milan Z. Bjelica and Nemanja Ignjatov and Mića Ćetković and Ivan Kaštelan},
url = {http://www.indel.etfbl.net/2014/resources/Proceedings_2014/INDEL_2014_Paper_41.pdf},
isbn = {978-99955-46-22-9},
year = {2014},
date = {2014-01-01},
booktitle = {International Symposium on Industrial Electronics},
pages = {248-252},
abstract = {The lack of easy to use environment to connect and manage the things is one of the obstacles that hold back multimodal personalized use of Internet of Things. We present a novel idea and a cloud framework with original elements that enable development of ecosystem for the user tailored deployment of things in the context of a personalized situational awareness. The benefits of the infrastructure are fortified with several use case scenarios that provide an insight to the platform’s innovative possibilities.},
howpublished = {M31},
note = {Keynote},
keywords = {cloud, context-aware, internet of things, keynote, smart homes},
pubstate = {published},
tppubtype = {conference}
}
The lack of easy to use environment to connect and manage the things is one of the obstacles that hold back multimodal personalized use of Internet of Things. We present a novel idea and a cloud framework with original elements that enable development of ecosystem for the user tailored deployment of things in the context of a personalized situational awareness. The benefits of the infrastructure are fortified with several use case scenarios that provide an insight to the platform’s innovative possibilities.
Integration of digital television in Android devices ConferenceKeynoteM32
Lunchtime Research Seminar, Communication & Computing Research Centre at Cultural, Communication and Computing Research Institute (C3RI), Room 921, Owen Building, Sheffield, 2012, (Keynote).
@conference{c3ri2,
title = {Integration of digital television in Android devices},
author = {Milan Z. Bjelica},
url = {http://www.milanbjelica.info/publications/c3ri_2012_invitation.pdf},
year = {2012},
date = {2012-02-17},
booktitle = {Lunchtime Research Seminar},
publisher = {Communication & Computing Research Centre at Cultural, Communication and Computing Research Institute (C3RI)},
address = {Room 921, Owen Building, Sheffield},
howpublished = {M32},
note = {Keynote},
keywords = {digital TV, keynote},
pubstate = {published},
tppubtype = {conference}
}
Context-aware lighting as an immersive user interface for mediating social interactions ConferenceKeynoteM31
Proceedings of the 10th WSEAS international conference on Applied computer and applied computational science, Venice, Italy, 2011, ISBN: 978-960-474-281-3, (Keynote).
@conference{wseas1,
title = {Context-aware lighting as an immersive user interface for mediating social interactions},
author = {Milan Z. Bjelica and Nikola Teslić},
url = {http://dl.acm.org/citation.cfm?id=1965653},
isbn = {978-960-474-281-3},
year = {2011},
date = {2011-03-08},
booktitle = {Proceedings of the 10th WSEAS international conference on Applied computer and applied computational science},
address = {Venice, Italy},
abstract = {Evolving generations of user interfaces are inadvertently seeking for revolution in the way people interact with computers and computer-based systems. We witness the birth of a potpourri of shape-shifting, translucent user interface forms or wearable gadgetry that is not likely to be adopted any time soon. Instead of proposing a next generation of the user interface, we argue that new means of interfacing should be sought for usability problems of marginal size, but immense impact to work and living. In this paper we propose an approach that addresses subtle usability issues of software applications and systems with regard to social interactions. Our approach introduces a context-aware platform that uses lighting effects programmed specifically for each state of context. We exercise the use of this platform in several important social interaction situations, such as to prevent interruptions or attract attention.},
howpublished = {M31},
note = {Keynote},
keywords = {availability estimation, context-aware, keynote, lighting control, user monitoring},
pubstate = {published},
tppubtype = {conference}
}
Evolving generations of user interfaces are inadvertently seeking for revolution in the way people interact with computers and computer-based systems. We witness the birth of a potpourri of shape-shifting, translucent user interface forms or wearable gadgetry that is not likely to be adopted any time soon. Instead of proposing a next generation of the user interface, we argue that new means of interfacing should be sought for usability problems of marginal size, but immense impact to work and living. In this paper we propose an approach that addresses subtle usability issues of software applications and systems with regard to social interactions. Our approach introduces a context-aware platform that uses lighting effects programmed specifically for each state of context. We exercise the use of this platform in several important social interaction situations, such as to prevent interruptions or attract attention.