Note: Visual AI Checkpoint is currently in Beta.
Visual AI Checkpoint enables instructors to verify that participants have achieved the expected states during a VM’s Guided Journey in both self-paced and instructor-led experiences.
One Visual AI Checkpoint can be added to each section of a Guided Journey as a visual validation step to ensure participants have the skills to perform the step properly and achieve the expected result.
CloudShare compares the participant’s current VM screen with a reference screenshot uploaded by the instructor. Only the areas selected by the instructor are compared.
This feature provides immediate pass/fail feedback and helps instructors identify where participants may be struggling.
How Visual AI Checkpoints Work
The feature verifies each participant's success in each section by comparing a reference screenshot pre-uploaded in the Editor with a participant’s live VM screen at the Visual AI Checkpoint stages of a Guided Journey.
- For each Visual AI Checkpoint, the instructor uploads a reference screenshot that shows the expected VM state. The instructor then marks the specific areas that must match the participant’s VM screen.
- When the participant clicks the Check button during a Guided Journey in the Participant Viewer, CloudShare takes a screenshot of the participant’s current VM screen and compares the marked areas against the reference screenshot. For example, it can check whether a certain Excel spreadsheet is open and whether a certain predefined area shows the same content.
- CloudShare’s Visual AI Checkpoint algorithm leverages computer vision to detect predefined visual elements and compare them against the expected results providing realtime insights into the successes and/or failures of each participant in a Guided Journey.
- If the marked areas match, the Checkpoint passes. If they do not match, the Checkpoint fails, and the participant may see a textual hint or visual marking indicating where the issue to be corrected is on the VM screen.
- Checkpoint results are saved as part of the participant’s progress, can be reviewed by instructors, and may be reported to a connected LMS when LMS reporting is configured.
Important: A failed Checkpoint does not block progress. Participants can continue to the next section at any time.