Latest events

AIML Research Showcase 2023

Dana Rawls — Fri, 28 Jun 2024 14:25:03 +0930

More than 100 people from across South Australian government, business and industry came together to get a rare insight into the latest in responsible artificial intelligence (AI) research.

AIML Lightning Talk: Engineering Intelligent Digital Systems

Dana Rawls — Fri, 28 Jun 2024 14:13:36 +0930

The Centre for Research on Engineering Software Technologies (CREST) team has been developing a trustworthy modularised Digital Twin Platform (DTP) that can be applied to autonomous and safety-critical cyber-physical systems (CPSs), such as smart factories, smart healthcare systems, and autonomous vehicles. The CREST DTP provides simulatable models of target systems (a.k.a digital twin models) based on runtime data streamed from real target systems.

AIML Mid-Week Seminar: Can biological neurons further improve our understanding of artificial neural networks and vice versa?

Dana Rawls — Fri, 28 Jun 2024 14:05:36 +0930

In the VPNL, we record from visual neurons in the dragonfly brain. A set of these neurons respond to optic-flow, encoding the dragonfly's own motion through the world. This underlies their remarkable ability to hover nearly stationary, then rapidly patrol at over 60KM/hour. Other neurons respond to tiny targets amidst cluttered surrounds, even predicting the target's future location following an occlusion. These target-detecting neurons can implement a winner-takes-all network, allowing for the selection of a single target amidst distractors (e.g. feasting on midges in a swarm).

AIML Mid-Week Seminars: What cat is that? and Machine learning for ecology: Using a CNN to understand the impacts of linear infrastructures on microhylid frogs in Papuea New Guinea

Dana Rawls — Fri, 28 Jun 2024 13:47:50 +0930

What cat is that? Victor Caquilpan's research project focused on using computer vision models to identify and track feral cats in camera trap images. Feral cats are a significant threat to the environment and accurately identifying them is crucial for conservation efforts. The study explores various methods including deep learning, feature extraction, and image processing to create an effective re-identification (Re-ID) model. The research systematically evaluates these models based on mAP and accuracy. By combining advanced computer vision and machine learning, this research aims to develop efficient solutions for feral cat RE-ID which can help monitor and manage feral cat populations and their impact on wildlife ecosystems.

AIML Mid-week Seminar: Using AI to predict clinical outcomes in dementia

Dana Rawls — Fri, 28 Jun 2024 13:13:10 +0930

In this presentation, Luke discussed the use of artificial intelligence and deep learning to analyse brain images in dementia, cognitively impaired and healthy individuals. The goal is to develop representations of disease progression that can differentiate normal aging from specific pathologies at both group and individual levels. The presentation aims to bridge the gap between group and individual disease progression analyses by combining morphometry, normative, and computational disease modelling techniques. It also outlines various approaches and results obtained so far, with plans to develop comprehensive normative models. This research has the potential to improve dementia severity assessment and guide precision medicine and aged care planning. Future work will explore different patient categories and intervention approaches.

AIML Special Presentation: Explaining the Uncertain - Stochastic Shapley values for Gaussian process models

Dana Rawls — Thu, 27 Jun 2024 14:34:46 +0930

In the rapidly evolving field of machine learning, it is important to quantify model uncertainty and explain algorithm decisions, especially for safety-critical domains such as healthcare. In this talk, Dr Chau presented a novel approach to explaining Gaussian processes which we term GP-SHAP. our method is based on the popular solution concept of Shapley values extended to stochastic cooperative games, resulting in explanations that are random variables.