Statistics Export

Overview

The Statistics Export plugin enables you to generate and download comprehensive reports containing essential annotation statistics from your project. These reports provide structured statistical insights from your project’s annotation data. It enables teams to analyze annotation distributions, identify patterns, and better understand dataset characteristics across assets, annotators, or custom attributes.

Statistics Export

Plugin Functionality

The Statistics Export Plugin processes project data and produces statistical summaries and visualizations based on the provided configuration.

• Provides annotation statistics at the project, asset, page (for video and multi-page assets) levels. This allows users to analyze both high-level trends across the dataset and detailed patterns within individual assets or frames.

• Supports grouping and aggregation of statistics based on a specified attribute.

• Computes several dataset-level metrics, including:

  • Number of pages/frames per asset, helping analyze the distribution of data across videos or multi-page documents

  • Unique object count, representing the number of distinct annotated objects across the dataset

• Generates visualizations for overall statistics, making it easier to interpret trends and insights without additional processing.

Running the Statistics Export plugin on your project generates multiple CSV files:

[Asset] File

Contains annotation statistics for each asset.

[Page] File

Provides annotation statistics at the page or frame level for each video assets, pdf documentations or multi-page assets.

[Stat] File

Includes overall statistics aggregated across the entire project, offering a high-level summary of annotation distribution and dataset composition.

[Page_Attribute] File

Breaks down page-level statistics based on the specified attribute in the Config JSON.

[Num_Page] File

Summarizes statistics based on the number of pages or frames per asset.

[Unique] File

Reports unique object counts across video frames, helping identify distinct objects throughout a video.

Supported Data Types

• Compatible with all data types available in AngoHub.

Supported Annotation Tools

• Compatible with all annotation tools available in AngoHub.

Plugin Configuration

From the Stage Filter field, pick the stages containing the tasks you'd like to get the export of. Similarly, from the Batch Filter field, you may select one or more batches the tasks of which will be exported. By default, all tasks are exported from all batches.

If you wish to receive an email when the export is complete, toggle Send Email on.

You may vary a number of settings related to your export from the Config JSON field. Each option is detailed below:

{
  "selected_external_ids": [],
  "additional_attribute_name": null,
  "include_key_frames_only": true,
  "logarithmic": false,
  "add_legend": true
}

• "selected_external_ids": Specifies a list of asset external IDs to include in the export. If provided, the plugin processes only the specified assets. If left empty, all assets in the selected batches and stages are included.

  • Example:

    • "selected_external_ids": ["12345", "12346"]

• "additional_attribute_name": Defines an optional attribute name used for grouping or aggregating statistics. This attribute must be defined as a classification under a tool.

  • Example:

    • "additional_attribute_name": null

    • "additional_attribute_name": "Color"

    • "additional_attribute_name": "Object ID"

• "include_key_frames_only": Specifies whether only key frames should be included in the metrics calculation. (For video assets only)

  • Example:

    • "include_key_frames_only": true

    • "include_key_frames_only": false

• "logarithmic": Determines whether a logarithmic scale is applied to the statistics in the generated visualization.

  • Example:

    • "logarithmic": true

    • "logarithmic": false

• "add_legend": Controls whether a legend is included in the generated visualization.

  • Example:

    • "add_legend": true

    • "add_legend": false