Generative Models for Generalised Skeleton-based Human Action Recognition. This project aims to develop innovative techniques rooted in generative models for more generalised human action recognition

Description

Generative Models for Generalised Skeleton-based Human Action Recognition. This project aims to develop innovative techniques rooted in generative models for more generalised human action recognition using privacy-preserving skeleton sequences. This project expects to contribute new knowledge in data-efficient learning, zero-shot learning, and domain adaptation through the development of novel methods. Expected outcomes of this project include novel techniques for generating skeleton data and enhancing action recognition models, enabling models to recognise unseen actions and adapt to diverse domains with limited training data. This should provide significant benefits to science, society, and the economy nationally and internationally, through various applications such as in autonomous vehicles and healthcare.. Scheme: Discovery Early Career Researcher Award. Field: 4611 - Machine Learning. Lead: Dr Qiuhong Ke