Creating Structured Layouts

Updated on Nov 3, 2024
Published on Sep 16, 2024
9 minute(s) read

If you plan on processing completed forms through Hyperscience, you first need to create a Structured layout for the form. After you add the layout to Hyperscience, you can use the Layout Editor to define what information the system should extract from the processed pages that match the specific layout.

Create a Structured layout

Go to Library > Layouts.
Click Add Layout.
Click Structured Layout, and then click Next.
Upload a PDF, TIFF, JPG, or PNG file in one of the following ways:
- Drag and drop the image file into the box provided.
- Click Choose File to find the file on your machine and upload it.
  - If you upload a multi-page TIFF or PDF file, a single layout variation will be created from the file.
If your form consists of multiple files, upload additional files by clicking Add More Files.
Click Next.
Enter a name for your layout in the Layout Name field.
Under Language, click on the drop-down list, and click on the language that you expect people will use when entering information in this layout’s documents.
Click Create.

By following these steps, you will create a layout with a single variation. You can add variations to the layout by following the steps in Adding a Variation to a Layout.

Access the Layout Editor

You can specify the information that should be extracted from submitted pages that match the layout by defining the field metadata in the Layout Editor.

To access the Layout Editor:

Go to Library > Layouts, and click on the name of the layout you want to edit.
Click on Layout Variations, and click on the name of the variation.

When in the Layout Editor, you’ll be making changes to the working version of the layout. To commit your changes—or to save without committing your changes—follow the steps in Saving and exiting below.

Using the Layout Editor

The layout’s title appears across the top of the page. The button to commit changes appears in the upper-right corner.

The Layout Editor for Structured layouts is split into three panels: page thumbnails, the main page view, and the field list.

Page viewer

To change between pages in the layout or layout variation, click on the page thumbnails on the left-most panel, or use the paging controls above the table of extracted fields and layout identifiers. Changing pages will automatically save your progress.

Layouts consist of fields. Each field specifies a piece of information to be extracted from a submitted page matched to that layout.

Field list

The right panel displays both Layout Identifiers and Fields together.

Fields are highlighted in blue. Layout IDs are highlighted in green.
Selecting a field or layout ID on the right panel will focus the page image in the middle panel to the corresponding bounding box for that field.
Use the search bar to find data types or field names for a field that has been defined in the layout.

When you create a Structured layout, the system attempts to identify its fields and adds them to the fields list. You can edit the bounding boxes and any settings for these fields.

Defining fields

On a layout for a Structured document, fields have two elements:

A bounding box that defines the area on the page where one expects to find a field.
Metadata, which consists of settings used by the machine to read and process the field value, as well as names used to label the information extracted from the field.

To define a field, click and drag on the page image to create a bounding box. Ensure the bounding box covers the area of the layout image where information is expected to appear.

Defining field metadata

After drawing the bounding box, define the field metadata in the field list on the right. The following pieces of metadata can be defined for each field on a Structured layout:

Field Name - labels the field throughout Hyperscience.

Data type - designates the type of data that is expected for the given field.

Output Name - allows you to provide a programmatic name for each field.

Transcription Supervision - allows you to specify the field’s transcription handling. The possible values are:

Autotranscribe - outputs the machine’s transcription value without generating manual transcription tasks.
Default - sends the field to Supervision as a manual transcription task if the machine’s confidence level on its transcription is below the threshold.
Always - guarantees manual review of the field.
Consensus - requires at least one manual transcription of the field regardless of the machine’s confidence.

For each field, a field name and data type must be defined. If either is missing, an error message saying "Missing Field Information" will be shown.

Additional Settings

There are additional settings and features to improve machine performance:

Multiline - enables the field for more than one line of text.
Dropout - indicates whether any background features in the field area should be included in the transcription.
Required - applies special logic to processing of submitted pages matched to that layout.

Editing multiple fields

You can select multiple fields from the field list to change parameters such as Data type, Supervision settings, Field configurations, and Size and location of bounding boxes.

Saving and exiting

All layout working versions auto-save:

To commit your changes, click Commit Changes.
To leave the Layout Editor without committing your changes, click the X in the upper-right corner.

To match submitted pages to a layout, the layout must have at least one committed version.