Inception: 09/01/2019 ~ 09/29/2019 Status: Completed

• Tasks:
  • Set up the camera AKDK in The Duchess of Kent Children’s Hospital at Sandy Bay to capture RGB-D images and label the landmarks on the RGB-D images.
  • Identify the project scope and finish a detailed project plan.
  • Build a project website.
  • Research on landmark detection projects.
• Milestones:
  • Equipment setup for data collection.
  • A detailed project plan.
  • A project website.

Elaboration: 09/30/2019 ~ 02/02/2020 Status: Completed

• Tasks:
  • Data preprocessing.
  • Data Augmentation.
  • Landmark detection on RGB-D images with HRNet.
  • Train the HRNet on CS GPU Farm.
• Milestones:
  • A customized dataset.
  • A deep learning model based on HRNet.
  • Designed experiments.

Construction: 02/03/2020 ~ 04/19/2020 Status: Completed

• Tasks:
  • Data preprocessing.
  • Data Augmentation.
  • X-ray synthesis from RGB-D images with the pix2pix model.
  • Train the pix2pix model on CS GPU Farm.
• Milestones:
  • A customized dataset.
  • A deep learning model based on the pix2pix model.
  • Designed experiments.