Skip to content
This repository was archived by the owner on Oct 31, 2023. It is now read-only.
This repository was archived by the owner on Oct 31, 2023. It is now read-only.

Training on custom dataset: Help with COCO skeleton #240

@ChristianIngwersen

Description

@ChristianIngwersen

Hi all,

I'm trying to train the model on a new custom dataset which uses the COCO joint format for both 2D and 3D joints (it's a multi view system where I have triangulated the 3D joints).

I'm a bit unsure how I should go about defining the Skeleton. I.e. in the Human3.6m dataset what are the parents in

h36m_skeleton = Skeleton(parents=[-1,  0,  1,  2,  3,  4,  0,  6,  7,  8,  9,  0, 11, 12, 13, 14, 12,
       16, 17, 18, 19, 20, 19, 22, 12, 24, 25, 26, 27, 28, 27, 30],
       joints_left=[6, 7, 8, 9, 10, 16, 17, 18, 19, 20, 21, 22, 23],
       joints_right=[1, 2, 3, 4, 5, 24, 25, 26, 27, 28, 29, 30, 31])

Based on the numbers and number of joints in human3.6m (32) i guess it is the connectivity, but I simply dont see the structure of the parents or how I can modify it to a COCO skeleton.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions