Feedback Aide supports AI-powered extraction of learner responses from uploaded images, enabling handwritten or printed work created on physical media to be converted into text for use in Feedback Aide sessions.

Feedback Aide natively supports HTML and plain text responses. To support physical learner submissions, Feedback Aide provides a text extraction utility through REST API utility endpoints that converts uploaded images into machine-readable text.

The extracted text can then be used as the learner response text when creating a Feedback Aide session. Optionally, uploaded images can also be stored within the Learnosity platform and attached to sessions for later display in the Feedback Aide session.

Text extraction capabilities

Text extraction is optimised for extracting learner response text only from uploaded images by making use of visual and contextual cues.

Supported features

• Extract handwritten learner content from images
• Extract printed learner content from images
• Upload and optionally store learner response images in the Learnosity platform
• Attach stored images to Feedback Aide sessions for UI display

Supported image formats

The following image formats are supported:

• PNG
• JPG / JPEG
• HEIC / HEIF

Known limitations

Current limitations are intentional while the feature continues to evolve to support broader workflows.

The following scenarios are not currently supported:

• Single responses spanning multiple pages
• Multiple responses spanning a single page
• Multiple responses spanning multiple pages

Prerequisites

In addition to some developer skills, ensure you have:

• An OAuth key and secret to generate a security token - see Feedback Aide Security Tokens
• Familiarity with how the Advanced - Feedback Aide RESTful Mode works

Generating a security token

Create a security token that includes the following scope:

text_uuid:<UUID4>

Replace <UUID4> with a randomly generated UUID4 value.

Example

text_uuid:2d66b27c-dde7-47d8-9238-7212fa7a74cc

The same UUID must later be used in the extraction endpoint URL.

Uploading an image for text extraction

Upload a supported image to the following endpoint:

/api/utility/extract-text/single-response/{text_uuid}

Replace {text_uuid} with the UUID used in the security token scope.

Example flow

1. Generate a UUID4 value.
2. Include the UUID in the security token scope.
3. Upload the learner response image to the extraction endpoint.
4. Receive extracted learner response text.
5. Use the extracted text as the response value when creating a Feedback Aide session in Getting Started with Feedback Aide API.

Displaying learner response images in sessions

Uploaded images can optionally be stored within the Learnosity platform for later display in Feedback Aide session interfaces.

When storing images:

• The uploaded image identifier should be included during session creation.
• Attached images can later be displayed alongside AI-generated feedback and extracted text.
  • If using createUi() method immediately after session creation, first call reload() method so the session generates an image download link.

Images do not need to be stored in Learnosity’s platform to appear in the Feedback Aide UI. Externally hosted images can also be displayed, provided they are publicly accessible and downloadable over the internet.

External images can be injected into the session UI using the createUi() handwritten_responses module.

For For more information, see Feedback Aide Session File Attachments.