Use Case Diagram

An overview of the use cases.

Name	U1 - Upload media
Description	The user uploads media for analysis.
Actors	User
Trigger	Upload media button on web page.
Steps	User visits the web page User fills in the upload form User presses the "Analyse Media" button
Exception Flow	Block the post and alert the user if form properties are not filled in.

Name	U2 - Add parameters
Description	The user adds analysis parameters.
Actors	User
Trigger	Add parameters button on web page.
Steps	User visits the web page User fills in desired parameters. User presses the "Add parameters" button
Exception Flow	Block the post and alert the user if form properties are not filled in.

Name	U3 - View job status
Description	The user views a currently ongoing analysis job.
Actors	User
Trigger	User clicks on the desired job on the job overview page.
Steps	User visits the web page User starts a job (U1) User visits the Job Overview page User clicks on the desired job
Exception Flow	The user is redirected to the job overview page if the job cannot be found.

Name	U4 - Get avatar reports
Description	The user views all reports/jobs of a single avatar.
Actors	User
Trigger	User selects an avatar out of the list of avatars
Steps	User visits the web page User visits the job overview page User selects the desired avatar
Exception Flow	The user is redirected to the job overview page if an error occurs.

There are two main use cases for the Avatar Pipeline.

The first and most common is general validation: users upload images or videos and want to validate whether the AUs therein are within tolerable values. This is what is displayed within the General Flow section. These parameters are hard-coded by the user and compared within an analysis module.

The second uses the same data but is retrieved differently: the user wants to be able to view results per avatar per test to "debug" an avatar (i.e. "give me every report with the avatar "Harry"). This is where the GitHub Actions-esque per-job overview comes into play. For this project, this was done using Azure Application Insights.