Labelbox Object Types

Updated 2 weeks ago by Florijan Stamenkovic

General Info

Field Types

Each Labelbox object field has an exact type. The types are as follows:

  • Int - a whole number (i.e. 42)
  • Float - a fractional number (i.e. 3.14)
  • String - a string of characters (i.e. "ProjectName")
  • Boolean - a true/false value
  • ID - Unique identifier, represented in Python as a String (i.e. "cjzmigz7muiji07944tvepkq5")
  • DateTime - a timestamp, represented in Python as a String (i.e. "2019-08-30T06:51:10.000Z")

Relationships

Most Labelbox objects are related to other objects. These relationships can be of the following types:

  • To-One. For example, a DataRow relates to exactly one Dataset.
  • To-Many. For example, a Dataset relates to multiple DataRows.

These relationships are combined into the following two-sided structures:

  • One-To-Many. For example, between a Dataset and DataRows.
  • Many-To-Many. For example, between Projects and Datasets.

Common Fields

Every Labelbox object has some common fields.

  • uid (ID) - unique identifier
  • created_at (DateTime) - automatically updated creation timestamp
  • updated_at (DateTime) - automatically updated last-modification timestamp

Labelbox Object Types

Project

  • name (String)
  • description (String)
  • setup_complete (DateTime)
  • datasets (To-Many Datasets)
  • created_by (To-One User)

Dataset

  • name (String)
  • projects (To-Many Projects)
  • data_rows (To-Many DataRows)

DataRow

  • external_id (String)
  • row_data (String)
  • dataset (To-One Dataset)

User

  • name (String)
  • nickname (String)
  • email (String)
  • organization (ToOne Organization)
  • created_tasks (ToMany Tasks)
  • projects (ToMany Projects)

Organization

  • name (String)
  • users (ToMany Users)

Task

  • name (String)
  • status (String)
  • completion_percentage (Float)


How did we do?