Fix/user mask overwrite

[aliFrancis]

Hotfix for issue I noticed during some testing. After annotating pixels, then training, the AI predictions would be incorrectly included in the user mask, and not only the final mask.

This PR fixes the problem by stopping the writing of those pixels to the newUserMaskData. In fact, newUserMaskData is not necessary at all in the legacyPredictMask() function, as predictions will never influence the user mask.

To reproduce error in code prior to hotfix:

(i) reset mask (remove all user and ai pixels)
(ii) draw pixels of 2 classes (observe the user mask correctly visualises the brushstrokes)
(iii)train AI model
(iv) observe user mask again, which will now be filled with all the AI's predictions, filling in the whole image around the user's annotations

[Copilot]

Pull request overview

Fixes a bug where AI-predicted pixels were being incorrectly marked as user-drawn pixels during legacyPredictMask(), causing the user mask view to be overwritten by predictions after training.

Changes:

• Stops writing AI prediction pixels into the user mask data during prediction application.
• Removes the unnecessary newUserMaskData copy and related post-update verification/logging.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[codecov-commenter]

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

✅ All modified and coverable lines are covered by tests.
❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.