VISION & LEARNING FOR AUTONOMOUS AI LAB

Research Fellow (Level A/B) - Monash University
Job No.: 645378
Employment type: Full-time
Location: Clayton campus - Monash University
Duration: 18-month fixed-term appointment (with the possibility of extention)
Remuneration: $75,115 - $101,944 pa Level A / $107,311 - $127,432 pa Level B

We invite motivated applicants to apply for one postdoctoral research position available on 3D vision, robot vision and robot perception. This project is a part of bigger project titled “HARNESS: Hierarchical Abstractions and Reasoning for Neuro-Symbolic Systems”. To achieve strong task performance in practice, autonomous systems, e.g. robots, need to provide high levels of robustness against uncertain and adversarial events, generalisability to unseen environments, and safety assurance. Purely data-driven, neural network-based approaches, however, suffer from fundamental limitations in achieving the above requirements, as they lack the ability to incorporate useful inductive biases in the form of symbolic background knowledge and to reason about them. This project proposes to deeply integrate symbolic modelling and reasoning techniques with neural network-based learning algorithms to holistically address the entire autonomy pipeline across all tasks from low-level perception to long-horizon planning. Our research program incorporates, invents and integrates techniques in neural networks and deep learning-based perception and prediction, efficient symbolic modelling and logic reasoning, causal modelling and inference, as well as deep hierarchical reinforcement learning. In this specific position, we are looking to hire a Research Fellow, working on Perception and prediction for an autonomous agent, e.g. a robot, using computer vision techniques.
Responsibility:

• Contribute to the project and help a large team of students and postdocs in both research & development
• Research and publish in the top-tier venues about the following topics:
  • Open-world 3D object detection
  • Open-world 3D object tracking
  • Object centric and sematic mapping
  • Intent/trajectory detection and prediction

We are looking for a highly motivated candidate, who is eager to get involved in cutting edge, creative research with real-world applications.

Essential Skills:

• A PhD degree in Computer Science, Mathematics, Physics, or Engineering.
• A background in machine learning and computer vision.
• Proficient programmer in one of the main Deep Learning libraries (e.g., Darknet, TensorFlow, PyTorch, Keras).
• Previous experience working with state-of-the-art computer vision or other machine learning/deep learning applications, especially related to 3D vision and robot vision.
• Fluency in written and spoken English, with an ability to communicate scientific ideas to an expert audience, both orally and written.

Track record of a publication in top tier Computer Vision, Robotics or Machine Learning conferences and/or journals would be also advantageous.

To apply for the position, please use this link. For any query, please send an email to Dr. Hamid Rezatofighi hamid.rezatofighi@monash.edu.