Pamuditha Somarathne

PhD Candidate | Human Motion Modelling | HCI, ML, Biosignal Processing

Generative Motion Modelling

← back to projects

Overview

This is my core PhD research at The University of Sydney: developing generative architectures for high-fidelity 3D human motion reconstruction. The work focuses on optimising latent space representations for sparse sensor data inputs, so that full-body motion (represented with SMPL) can be reconstructed from limited sensing.

Vision

A core driver of this research is improving the accessibility of emerging technologies: predictive models of human motion can make gesture systems more adaptable and equitable, creating more intuitive and inclusive digital experiences.

Generative AI Human Motion SMPL PyTorch HCI