Prof. Konstantinos Alexis

Title of talk: "Resilient Autonomy through Multimodal Perception”

Abstract: State-of-the-art autonomy methods remain fragmented with controllers, sensor fusion pipelines, and learning algorithms typically tailored to narrow robot morphologies and operating regimes. This specialization has historically been necessary to achieve operational results, but inevitably limits generalization and slows the pace of innovation. A common blueprint for autonomy is necessary. This talk outlines a vision toward Unified Resilient Autonomy that is applicable across diverse robot configurations, whether flying, aquatic, or ground systems. By pursuing a common autonomy architecture and leveraging the lessons learned from its broad evaluation in extreme conditions, we demonstrate resilient functionality that transfers across embodiments. The discussion will highlight both the underlying methods that enable this unification as well as concrete results from field testing in unconventional environments - such as subterranean settings, ship ballast tanks, and submarine bunkers.

Bio: Prof. Dr. Kostas Alexis conducts research in the domain of resilient robotic autonomy. Through his studies he has examined the potential of autonomous systems to navigate extreme environments by presenting resourcefulness, robustness and redundancy. His research areas include the domains of robot control, path planning, and Simultaneous Localization And Mapping, while across these disciplines a holistic approach is taken to facilitate enhanced resilient as demonstrated through field robotics. He has been the PI in major international grants sucha s the DARPA Subterranean Challenge and his funding portfolio includes grants from US sources (e.g., DARPA, NSF, DOE, USDA), EU (e.g., Horizon 2020 programs), as well as the Norwegian Research Council.

Prof. Hedvig Kjellstrøm

Title of talk: "Building computer models of human cognition and behavior"

Abstract: In the last decade, development of deep-learning methods for reconstructing, tracking and synthesizing human motion and facial expression in video has come a long way, and this technology is quite mature and is used for personal video editing, gaming, visualization of football games and much more. Even more recently, there has also been an explosive development of large language models as we all know, and there is also a large research focus on models that analyze language and visual cues.

Building on this development, the work in my lab focuses on modeling the underlying mechanisms behind human non-verbal and verbal behavior, from a cognitive and behavioral perspective. In this talk, I will present a number of projects addressing this from different viewpoints, for conversational gesture synthesis, equine pain detection, and emotion-centered facial expression recognition

Bio: Hedvig Kjellström is a Professor in the Department of Robotics, Perception and Learning at KTH Royal Institute of Technology, Sweden, and also affiliated with Swedish e-Science Research Centre and Max Planck Institute for Intelligent Systems, Germany. She received an MSc in Engineering Physics and a PhD in Computer Science from KTH in 1997 and 2001, respectively, and thereafter worked at the Swedish Defence Research Agency, before returning to a faculty position at KTH. Her present research focuses on methods for enabling artificial agents to interpret human and animal behavior. These ideas are applied in the study of human aesthetic bodily expressions such as in music and dance, modeling and interpreting human communicative behavior, and the understanding of animal behavior and experiences. In order to accomplish this, methods are developed for agents to perceive the world and build representations of it through vision.

Hedvig has received several prizes for her research, including the 2010 Koenderink Prize for fundamental contributions in computer vision. She has written around 150 papers in the fields of computer vision, machine learning, robotics, information fusion, cognitive science, speech, and human-computer interaction. She is mostly active within computer vision, where she is an Editor-in-Chief for CVIU, a Program Chair for CVPR 2025, and regularly serves as Area Chair for the major conferences.

Dr. Paolo Braca

Title of talk: "From Predictive Brains to Predictive Systems: Situational Awareness in the Era of AI. – Uncertainty, Scale, and Trust: High-Dimensional Statistics and Maritime Applications”

Abstract: The human brain does not simply observe the world—it continuously predicts it. Perception emerges from the interaction between internal models and sensory evidence, a process described as a “controlled hallucination” by Anil Seth. Interestingly, this perspective resonates with established models of situational awareness, such as the Endsley framework, as well as with their command-and-control counterparts, including the OODA loop and the F2T2EA decision cycle.

In this talk, I will discuss recent advances in artificial intelligence and information fusion that are transforming situational awareness systems into large-scale predictive systems capable of anticipating complex operational environments. Drawing on my experience in multi-sensor fusion, statistical learning, and signal processing, I will present examples from the open literature ranging from vessel trajectory prediction and anomaly detection to recent work on high-dimensional statistics and large deviations theory for understanding and controlling the performance of modern machine learning systems.

I will conclude by discussing the fundamental challenges that arise when deploying AI-driven situational awareness systems in real-world environments—particularly those related to uncertainty, scalability, and trustworthiness—and how advances in information fusion may help address them in future surveillance and decision-support systems.

Bio: Dr. Paolo Braca is an IEEE Fellow and has been with the NATO Science and Technology Organization (STO) Centre for Maritime Research and Experimentation (CMRE) since 2011. His research focuses on machine learning, statistical signal processing, information fusion, and multi-sensor multi-target tracking, with applications to maritime surveillance, autonomous systems, and decision-support technologies. He has led research projects funded by NATO, the European Commission, the US Office of Naval Research, and other institutions. He has authored more than 200 publications in international journals, conferences, and NATO technical reports. Dr. Braca currently serves as Editor-in-Chief of the ISIF Journal of Advances in Information Fusion, and as Associate Editor for IEEE Transactions on Aerospace and Electronic Systems and IET Radar, Sonar & Navigation. He previously served as Associate Editor for IEEE Signal Processing Magazine, IEEE Transactions on Signal Processing, and the EURASIP Journal on Advances in Signal Processing. He received the NATO STO Scientific Achievement Award in 2017, the NATO STO Excellence Award (team leader) in 2020, and the IET Premium Award in 2019. He is listed among Stanford University’s World’s Top 2% Scientists. Dr. Braca holds the Italian National Scientific Qualification to serve as Full and Associate Professor in the fields of Telecommunications, and Information Processing Systems.