So How Likely Is a Car Crash? Estimating Rare Failure Events in Simulated Scenarios

Date: 
Thu, 24/10/2024 - 13:00
Location: 
Informatics Forum. G.07
Speaker: 
Craig Innes

Abstract: Scenario-based Simulated Testing is a common paradigm for testing complex cyber-physical systems like autonomous vehicles. This is particularly true for systems containing black-box ML components, which make formal verification or traditional engineering analysis difficult. However, answering the simple question of "how likely is it that the system will fail to meet its specification?" can actually be a significant challenge, especially when the true failure probability is rare.

Break Free & Build: The PhD-to-Tech-Founder Journey

Date: 
Fri, 25/10/2024 - 15:30 to 16:30
Location: 
Informatics forum, G.03
Speaker: 
Rolf Baxter
Oosto

PhD students - Are you curious about what lies beyond the academic path?
Join us for a casual, inspiring session with Informatics alumnus and serial tech founder Rolf Baxter, as he shares his journey from PhD student to building billion-dollar companies. In just a decade, Rolf has built Oosto, a leading computer vision company, launched and sold Zenith AI—a no-code AI platform—and is now on his third tech venture.

Robotic Surgery Technologies Workshop

Date: 
Fri, 11/10/2024 - 10:00 to 12:00
Location: 
Room 4.40 (Mini Forum 2) at the Informatics Forum, Newington, Edinburgh, EH8 9AB
Speaker: 
Multiple (Mustafa Suphi Erden, Mohsen Khadem)
National Robotarium at Heriot-Watt University, in partnership with the School of Informatics at the University of Edinburgh

The National Robotarium at Heriot-Watt University, in partnership with the School of Informatics at the University of Edinburgh, is organizing a Robotic Surgery Technologies Workshop. This event aims to promote collaboration between researchers across Edinburgh, bringing together experts and individuals with a shared interest in advancing robotic surgery technologies.

Optimal control and learning for contact-rich robotics

Date: 
Wed, 09/10/2024 - 15:00 to 16:00
Location: 
Informatics, G.03
Speaker: 
Dr Majid Khadiv
University of Munich

Abstract: The past few years has witnessed significant progress in the field of legged locomotion and manipulation. This is mainly due to the availability of high-performance hardware as well as development of algorithms that scale to high-dimensional, hybrid and under-actuated systems. In this talk, I will present my recent research efforts, mainly on the algorithmic side, on developing efficient predictive controllers that can be complemented with supervised/reinforcement learning for real-time execution in the real world.

IPAB Workshop Video Understanding with Limited Resources

Date: 
Thu, 26/09/2024 - 13:00
Location: 
Informatics Forum. G.03
Speaker: 
Laura Sevilla-Lara
University of Edinburgh

Abstract: Understanding video is a fundamental skill of intelligent systems such as robots or virtual assistants, to be able to move and interact with the world or with humans. However, learning how to understand video is very expensive, both in terms of data and compute. This cost has a wide range of effects, from carbon emissions to holding back progress in research. In this talk I will discuss the solutions that we are designing in my lab, as well as the underlying principles and takeaway messages for future work in this area.

Adaptive Experience Sharing in Multi-Agent Reinforcement Learning: A Unified Approach for Heterogeneous and Homogeneous Agents

Date: 
Thu, 19/09/2024 - 13:00 to 14:00
Location: 
Informatics Forum, G.03
Speaker: 
Atish Dixit
University of Edinburgh

(IPAB Workshop): Adaptive Experience Sharing in Multi-Agent Reinforcement Learning: A Unified Approach for Heterogeneous and Homogeneous Agents