Career Opportunities

Career Opportunities

Help us shape AI’s future – and we’ll transform yours.

Join the largest university machine learning group in Australia and play a pivotal role in the new Centre for Augmented Reasoning research projects.

The Centre for Augmented Reasoning (CAR) is a vital new hub for Australia’s high-calibre machine learning expertise. Part of the Australian Institute for Machine Learning (AIML), the centre unites people and research to drive forward our nation’s participation in the machine learning-enabled global economy.

Established in 2021 through investment by the Department of Education, Skills and Employment, our mission is to create computers that understand human instructions and needs through natural conversation. Our vision is for the University of Adelaide to be recognised globally as a centre of research excellence in machine learning and augmented reasoning by 2026.


view of AIML building on north terrace adelaide

Next generation machine learning

We’re improving the underlying tools that enable more functional AI to deliver solutions in the real world.

We're seeking four postdoctoral fellows in the following projects.

1.1 Data-Efficient Learning for Integrated Scene Understanding

Researchers: Dr Lingqiao Liu, Professor Chunhua Shen, and Professor Anton van den Hengel

Seeking one postdoctoral fellow.

You will create integrated scene understanding systems, which are able to recognise as many detailed visual concepts as possible in a unified framework from a given image. Within research theme Next Generation Machine Learning, fundamental algorithms for image classification, object detection, optical flow estimation, semantic and instance segmentation etc. will be integrated into an overall framework to achieve holistic, integrated and complete understanding of visual contents, termed ‘Integrated Scene Understanding’ to narrow the gap between computer systems and human vision systems in terms of the depth of understanding.

The aim is to:

  • unify many fundamental vision tasks into a common deep learning framework
  • develop deep learning algorithms that do not need a large amount of data with detailed annotation for computer vision applications, i.e., deep learning with imperfect data.

Selection Criteria  apply

1.2 Deep Learning at Scale

Researchers: Dr Lingqiao Liu, Professor Chunhua Shen, and Professor Anton van den Hengel

Seeking one postdoctoral fellow.

You will investigate fundamental bottlenecks when attempting to develop deep learning applications at scale. Within research theme Next Generation Machine Learning, your research will overcome two primary limitations of deep learning and will greatly increase its domain of practical application.

Fundamental limitations to developing deep learning applications at scale will be addressed to increase its domain of practical application and enhance a neural network model. The aim is to:

  • design compact, or binary/quantised networks, enabling large-scale deployment of deep learning to mobile devices
  • develop efficient neural architecture search methods that are orders of magnitude faster than previously reported, abstracting away the most complex part of deep learning.

Selection criteria  apply

1.3 Next Generation Causality

Researcher: Professor Javen Shi

Seeking one postdoctoral fellow.

You will revisit, reimagine, rejuvenate causality, to build next generation causality capability on some of the following topics:

  • higher order non-linear causal relation discovery;
  • optimal intervention/experimental trials by active learning;
  • causal meta learning. For example, different growers have different sensors in place to monitor their plants, and often at varying rates as well. The variables available change significantly across different sites, growers, and plants. Meta learning techniques will be investigated to build multiple causal models (e.g. one causal model per site), which collectively will improve the master/meta model, and in turn, improve each causal models as well.
  • causal reinforcement learning (RL) to explore how to build a more powerful model-based RL using causality to describe the environment and to generate counterfactual episodes, and enable the agent to imagine.
  • instrumental variables to avoid potential catastrophic consequences caused by directly intervening variables in complex and unsure causal structures;
  • other fundamental causality areas.

Selection criteria  apply

1.4 Quantum Augmented Reasoning (QuAR)

Researcher: Professor Tat-Jun Chin

Seeking one postdoctoral fellow.

You will investigate quantum computing algorithms that could power the machine learning algorithms that underpin augmented reasoning. Within the research theme Next Generation Machine Learning, the research will involve the development of quantum machine learning algorithms, experimentation on quantum computers, and the development of augmented reasoning applications.

The candidate should have a strong background in machine learning, physics and mathematical programming.

Selection criteria  apply

Interactive machine learning

We’re developing new tools that improve the ability of humans and machines to interact more effectively and efficiently.

We're seeking eight postdoctoral fellows in the following projects.


2.1 Continual Learning with Modulated Memory Augmentation

Researcher: Professor Anton van den Hengel

Seeking one postdoctoral fellow.

You will investigate continual learning (CL) for the next-generation machine learning models that can learn continually while interacting with the world via the approaches (but not restricted to):

  • memory-augmented neural networks
  • meta-learning
  • self-supervised learning
  • unsupervised learning
  • causal inference

Within research theme Interactive Machine Learning, the aim is to resolve CL via designing new deep neural networks (NNs) with balanced plasticity and stability while learning on continually varying tasks. The models should quickly handle the new data/tasks, manage the new skills, and keep the performances on the old tasks.

Selection criteria  apply

2.2 Learning to Interact and Reason in Open-Ended Tasks

Researcher: Dr Ehsan Abbasnejad

Seeking one postdoctoral fellow.

You will create a new area of research in vision-and-language tasks called (potentially open-ended) open-domain visual dialogue (ODVD). Motivated by the need in improved reasoning in conversational agents for dialogue systems causal and counterfactual reasoning in ODVD will be employed. The purpose of ODVD is to extract what is the intent of humans, change human actions and learn to act indistinguishably humane. Potential tasks include:

  • creating New Benchmarks for ODVD, which may include 1) An autonomous driving dialogue where the agent converses about where to go, speed, route, places to stop or park, etc. and 2) An embodied agent with navigation manipulation that conducts dialogue with the agent about the desired outcome
  • develop Dialogue as Interventions in ODVD. Human intent and its potential outcome is modelled. It is important to consider what could humans do differently. Alternatively what agents could utter differently to intervene in human’s actions or response
  • Research in Continual Imitation Learning where the agent learns from the demonstration dialogues with humans.

Selection criteria  apply

2.4 Augmented Reasoning in Vision-and-Language

Researcher: Dr Qi Wu

Seeking three postdoctoral fellows.

You will research and develop novel algorithms that can ground vision language to real-world daily applications.

Within Research Theme 2: Interactive Machine Learning, your research in Computer Vision, Visual Question Answering, and Vision-and-Language technology more broadly will develop and explore five research themes:

  • integrating prior knowledge into vision-and-language with symbolic reasoning
  • Explainable Visual Question Answering
  • Vision-and-Language in the wild: causality, reliability and generalisability
  • fine-grained control of image2text and text2image generation
  • Human-centred Vision-and-Language Navigation.

Selection criteria apply

2.5 Embodied Visual AI

Researcher: Professor Ian Reid

Seeking two postdoctoral fellows.

You will investigate visual and geometric scene understanding as a means to interact with, query and learn about the environment, and embodied visual AI in which scene understanding is deployed in physical autonomous systems. The project aims to develop:

  • new or improved ways of learning for embodied systems especially through deliberative interaction
  • representations of the world, and ways to use the world “as its own best memory”
  • the best ways to combine the power of data-driven learning with known tasks.

You will research the limitations of, extensions to teach-and-repeat and reinforcement learning, and how these interact with simulation (and the transferability of sim to real).

Selection criteria apply

2.7 Art Intelligence Agency

Researchers: Professor Anton van den HengelDr Jamie Sherrah and Professor Thomas Hajdu

Seeking one postdoctoral fellow.

The postdoctoral researcher for the Art Intelligence Agency will work with world-renowned musicians and artists to develop new ideas, algorithms and creations at the frontier of music, art and AI. Working with world-class researchers in machine learning at AIML, you will publish impactful papers in deep learning areas such as music and voice generation and natural language models. 

This is an opportunity to firmly establish your research career working alongside some of the world’s best academics in machine learning and computer vision, and also work with world famous artists.

Selection criteria apply

Knowledge, representation and generalisation

We’re developing new models for reasoning and storing learned information and rules that can be called upon by machines to solve new challenges.

We're seeking two postdoctoral fellows in the following projects.


3.1 Scalable Compositional Semantic Meaning Representation and Reasoning

Researcher: Dr Lingqiao Liu

Seeking one postdoctoral fellow

You will work on developing novel semantic representations for natural languages. Specifically, this project within Research Theme 3: Knowledge, Representation and Generalisation aims to develop natural language understanding methods that can generalise across scenarios and support complex reasoning.

Selection criteria apply

3.2 Towards "Small Data, Big Problem" Learning and Reasoning Paradigms

Researcher: Dr Lingqiao Liu

Seeking one postdoctoral fellow

You will work on developing novel machine learning and reasoning paradigms. Specifically, the project within Research Theme 3: Knowledge, Representation and Generalisation aims to develop solutions for "small data, big problem" scenarios, with the focus on improving the generalisability of machine learning systems across domains and tasks.

Selection criteria apply

Machine learning driven science discovery

Our goal is to apply the capabilities of CAR to solving high-value, real-world problems that currently are difficult to solve using traditional approaches. We aim to convert machine learning capability into solutions for intractable challenges faced by humans now and in the future.

We're seeking two postdoctoral fellows in the following projects.


4.2 Building Causal Models for Predicting Treatment Outcome in Patients - Towards the Automated Bioinformatician

Researchers: Dr Johan Verjans, Mohsen Dorraki and Professor Javen Shi

Seeking one postdoctoral fellow.

You will work to achieve the automation of core capabilities of a bioinformatician, which could lead to a real-time data and patient characterisation platform for researchers and clinicians. This position, within Research Theme 4: Machine Learning Driven Science Discovery, also has the responsibility to facilitate collaboration in the field of -omics driven patient diagnosis and prognosis, drug efficacy and toxicity prediction, etc., enabling CAR into future commercial partnerships with pharma companies in diagnostic and drug development.  

In this position you will develop, implement and test robust machine learning and computer vision algorithms for solving important questions in biomedical and clinical research using multi-omics data and other information modalities (including imaging), publishing in conferences and journals, and participation and leadership in machine learning workshops and presentations.  

Selection criteria apply

4.3 AI-driven Discovery for Energy Materials

Researchers: Professor Shizhang Qiao and Professor Javen Shi

Seeking two postdoctoral fellows.

You will work with domain experts (such as Laureate Fellow Professor Qiao) on the machine learning aspect of the project to:

  • develop machine learning models to learn mapping from the input ingredient and environment conditions to the final output material from the observed/collected data, potentially using active learning to suggest trials and counterfactual reasoning to learn from unobserved events 
  • develop novel deep reinforcement learning methods to exploit and explore best suitable formulas for material with desirable properties
  • apply to accelerate the discovery for new energy in both targeted experimental synthesis and theoretical simulation on catalysis and battery

Selection criteria apply

A quality-of-life ‘singularity’

Living and working in Adelaide, South Australia offers a remarkable quality of life. Ample sunshine, parklands and natural beauty. Uncongested roads and affordable modern living. Access to state-of-the-art technology and world-class industry. Fine dining, elite sport and a vibrant cultural life.

Life in Adelaide Workplace benefits About our uni