Skip to main content

Summary

Clicking on a project in the Projects view takes you to its 'Project dashboard'. The first screen you see is a summary of your project's statistics.

The statistics are broken down into 3 categories:

  • Productivity: a summary of your labeling activity
  • Label statistics: how your labels are distributed across datasets, objects and classifications
  • Instance statistics: how a particular class is distributed across datasets

Definitions of classes, instances and labels#

  • Class: Part of the project's ontology. Creates a definition for what an instance of the class can contain.
    • For example the ontology of a project annotating traffic videos would have classes Car, Truck, Bicycle etc
  • Instance: Occurrence of a class.
    • For example Car (0) is an instance of the Car class
  • Label: An annotation of an instance on a specific frame.
    • For example the annotation of Car (0) on frame 201 is a label

Productivity#

This is a summary of all the labeling activity in your project. The top graph shows the number of labels added, edited or deleted, while the lower graph shows the distribution of time taken for these activities. A pie chart at the bottom shows your collaborator breakdown when viewing 'Total' productivity.

  • Time frame: use the dropdown on the left to view the productivity across different time frames
  • Activity type: use the buttons in the middle to switch between adds, edits and deletes
  • Team breakdown: use the buttons on the right to switch between different collaborators

Label statistics#

This is a summary of how your labels are distributed. The pie chart shows a breakdown of how many labels there are for a given class.

  • Dataset: use the dropdown on the left to view this breakdown across different datasets
  • Class type: use the buttons on the right to see a breakdown between objects and classifications
  • Attributes: click on a pie chart segment to see the attributes for that class

Instance statistics#

This is a summary of how a given class is distributed across your project's datasets. The pie chart shows a breakdown of how that class is split across the datasets.

  • Class: use the dropdown on the left to select a particular class
  • Frame: use the toggle on the right to switch between frame numbers and timestamps for the listed labels