Human Trajectory/Body Motion Forecasting from Visual sensors

Primary supervisor

Hamid Rezatofighi

The ability to forecast human trajectory and/or body motion (i.e. pose dynamics and trajectory) from camera or other visual sensors is an essential component for many real-world applications, including robotics, healthcare, detection of perilous behavioural patterns in surveillance systems. However, this problem is very challenging; because there could potentially exist several valid possibilities for a future human body motion in many similar situations and human motion is naturally influenced by the context and the component of the scene/ environment and the other people's behaviour and activities. In this project, we aim to develop such a physically and socially plausible framework for this problem.

Student cohort

Double Semester

URLs/references

https://vl4ai.erc.monash.edu/research.html

https://arxiv.org/pdf/2104.04029.pdf

https://arxiv.org/pdf/2007.06843.pdf

https://papers.nips.cc/paper/2019/file/d09bf41544a3365a46c9077ebb5e35c3-Paper.pdf

https://openaccess.thecvf.com/content_CVPR_2019/papers/Sadeghian_SoPhie_An_Attentive_GAN_for_Predicting_Paths_Compliant_to_Social_CVPR_2019_paper.pdf

Required knowledge

  1. Good coding skills in a variety of coding languages
  2. Previous experience working with deep learning models for different tasks
  3. ​​​​​Proficient programming skills in Python and one of the main deep learning libraries (e.g., TensorFlow, PyTorch, Keras)