Towards Data-Efficient Future Action Prediction in the Wild

  • This project aims to build state-of-the-art deep learning models to predict future actions in videos with a handful of labeled examples. The project expects to produce the next great step for machine intelligence - the potential to explore a handful of labeled examples to better understand, interpret and infer human actions. Expected outcomes of this project lay theoretical foundations for learning future action prediction in the wild scenario and build the next generation of intelligent systems to accommodate limited supervision. This should benefit science, society, and the economy nationally through the applications of autonomous vehicles, sensor technologies, and cybersecurity.

Australian Research Council Discovery Early Career Researcher Award (DECRA), Awarded in November 2018.

A Machine Learning Framework for Concrete Workability Estimation

  • Develop a vision-based method for monitoring concrete's workability in the mixing bowl that is more accurate than other non-vision-based approaches.
  • Improve workability predictions by integrating multimodal data, i.e., machine vision and machine learning, with other sensors on agitator trucks, e.g., hydraulic pressure and rotation measurements.
  • Validate the effectiveness of developed technologies in practice through extensive field trials.

Australian Research Council Linkage Projects, granted in 2023

An intelligent condition-monitoring system for mineral screening machines

  • Develop a dynamic model to capture the typical vibration features of vibrating screens;
  • Develop new vibration-based techniques for fault diagnostics of bearings and gears, by combining signal processing techniques and source separation algorithms to extract weak fault signals;
  • Develop a fracture mechanics-based model to allow detailed root cause analysis and to assist the rational prediction of remaining useful life of gears;
  • Develop a digital twin-based framework by integrating physics model-based approaches and datadriven approaches using real-time measured data to achieve prognostics-based maintenance decision-making;
  • Based on field measurement and validation of diagnostic and prognostic algorithms, develop an intelligent condition-monitoring system for screening machines.

Australian Research Council Linkage Projects, granted in 2023