Join us at the next MMM & MAC. First up are PhD students: Sylvain Daronnat and then Matt Cabanag on their projects about Human-Autonomy Teaming (HAT). At the MAC we have Professor Robert Mahony on Event Camera Processing
NOTE: Both sessions will take place via Zoom ID 986 9057 6845.
12pm - MARCS Monday Meeting
Topic: Trust, Cognitive Load and Performance in Human-Agent Collaboration
Speaker: Sylvain Daronnat - PhD student, University of Strathclyde, Glasgow
Abstract: Collaborative agents help human operators complete tasks more efficiently in situations where they act as decision-aid systems or "virtual teammates". Trust is a prerequisite for this collaboration to work. In this talk, Sylvain will be going over two studies where analysis was done on how the level of reliability and the types of errors an agent make can influence trust, reliance, task performance and cognitive load. These two experiments were conducted via a simple game-like framework that enables the study and evolution of the human-agent relationship over time.
Biograpy: Sylvain Daronnat is PhD student in Computer and Information Sciences at the University of Strathclyde in Glasgow. His research interests lie in HCI and Human Factors. He creates interactive environments where users interact with different virtual agents, recording users' interactions through qualitative and quantitative means. He uses data science tools to create more effective automated agents, capable of adapting to changes in users' behaviours. He is also interested in the field of NLP, where he did his Master's degree, with a particular interest in NLG.
Topic: A Model of Trust in Human Autonomy Teaming (HAT) Based on Predictability and Shared Mental Models
Speaker: Matt Cabanag - PhD student, MARCS Institute, Western Sydney University
Abstract: Trust in autonomous teammates had previously been expressed in various ways, including DeVisser's transactional model of trust violation and repair. Matt will outline an alternative model whereby trust in autonomy is viewed as a result of beneficial and/or predictable behaviour. In this paradigm, predictability is enabled by shared mental models between teammates, which in turn is facilitated by cognitively efficient agent transparency. Matt will also give an update on building online experiments in the era of COVID-19 restrictions.
1pm – MARCS Afternoon Colloquium
Topic: Asynchronous Linear Filters for Event Camera Processing
Speakers: Professor Robert Mahony
Abstract: This talk presents an overview of recent work in image state reconstruction for event cameras. Events are modelled as a continuous-time signal comprising a sequence of dirac delta functions. Image reconstruction is conceptualised and solved as a continuous-time filter design. By solving for the linear response of the filter equations explicitly we derive an asynchronous update equation that is simple, robust and very low computational complexity. The approach can be applied to a range of image processing tasks including image reconstruction, image gradient computation, image filtering and computation and tracking of
Biography: Robert Mahony is a Professor in the Research School of Engineering at the Australian National University. He received a PhD in 1995 (systems engineering) and a BSc in 1989 (applied mathematics and geology) both from the Australian National University. He worked as a marine seismic geophysicist and an industrial research scientist before completing a postdoctoral fellowship in France and a Logan Fellowship at Monash University in Australia. He has held his post at ANU since 2001. His research interests are in non-linear control theory with applications in robotics, geometric optimisation techniques and systems theory.
We thank you for your continued attendance at the MMM's and look forward to seeing you again.
MARCS staff and students are reminded that all meetings and workshops have an important role in building and maintaining the sense of community which is central to the success of MARCS as a cooperative and energetic research institute. Your attendance is both welcomed and expected.
The zoom ID is: 986 9057 6845 . Link to zoom https://uws.zoom.us/j/98690576845