3D Clean-Up of 3D Face Dataset for 5G Holographic Communication

 

Défi de développement technologique 5G (ENCQOR) pour PME

3D Clean-Up of 3D Face Dataset for 5G Holographic Communication

Challenge Launch Date:

April 20, 2020

Challenge Deadline: 

May 31, 2020

Challenge Statement:

For each subject in our 3D Face dataset, align and track a provided 3D facial “template” mesh through consecutive frames of high-quality 3D scan data.

This is the first of several data collection/curation/infrastructure challenges in service of our 5G-powered holographic communication pipeline.

Project Partner:

Ericsson Canada Inc.

Timeline:

Approximately 6 months with delivery date in 2020

Available Funding:

Up to 200 000$ CAD

Applicant Type:

SME with less 500 employees registered in Quebec

Location:

At SME location in Quebec with 5G testing -if/as needed- in one of ENCQOR 5G Hubs in MTL or QC.

Project Details:

There are 25+ subjects in the dataset, with 5000+ consecutive frames of high-quality 3D scan data per subject.

Each subject has been recorded hitting a common set of facial expressions from the neutral pose, in the same order.

To make the tracking process easier for challengers, the template mesh will be provided in a variety of “initial” deformations—one for each facial expression the subjects were directed to hit.

Additionally, software licenses to R3DS’s Wrap3 will be provided, if desired. Wrap3 can be used to automatically (but roughly) deform a template mesh to the 3D scan data from each frame; the artists’ task will then be to make sure this deformation as correct as possible. With such an approach, we expect for a throughput of 20 frames / hr.

Challenge can be broken up into chunks, for multiple challengers to participate in parallel. This may be necessary to hit the 2 month turnaround for the entire dataset.

Written Q&A will be supported. Please submit your written questions to  aauger@promptinnov.com by May 1st for first batch and May 15th for the second. All questions and answers will be available to all challenge participants

Project Goals/Outcomes:

A high-fidelity 3D Face database which is “complete” across identity, expression, dialogue, and emotion; in service of our 5G-powered holographic communication pipeline.

Applicant Capabilities:

Previous experience with any 3D artist toolset.

Additional Information:

  • Interested SMEs can submit a (max) 10 pages document (in .pdf) with the Statement of interest describing its expertise and capabilities using the form below.
  • Selected SMEs will be invited for a live demo/presentation with the project owners
  • It is highly recommended that interested SMEs participate to one of the free 5G trainings offered by ENCQOR (https://quebec.encqor.ca/formations/)

Subscribe to our newsletter

Join our mailing list to receive the latest news and updates from ENCQOR 5G.

Thank you for your subscription!