Review what the AI detected

更新於 2026年7月22日

Auto detection covers faces, people, vehicles, and license plates. This article shows how to review the results and fill any gaps.

The target list

Each detection appears as a separate target with its own thumbnail. Clicking a target highlights it on the canvas, and clicking a detection box on the canvas selects it in the list, so you can work in whichever direction is faster.

Mask or keep

Toggle each target between mask and keep. A common workflow for group photos: leave everyone masked, then switch only the people who gave consent to keep.

When something is missed

  • Click it: point at the object on the canvas and it becomes a target. This works for things detection does not look for, like a phone screen or a name tag.
  • Draw a region: for full control, draw a manual box anywhere. Manual regions behave like any other target.

When something is wrong

A false positive, such as a poster face, is harmless: switch it to keep and it stays untouched in the export.

這篇文章有幫助嗎?

相關文章

還需要幫助?

我們的支援團隊隨時為您服務,通常在一個工作日內回覆。

聯絡支援