P: Adjustable Accuracy Levels for Face Recognition in Lightroom Classic

Lightroom Classic’s Face Recognition currently attempts to detect all faces in an image, including very small, blurred, distant, or partially visible faces. In photos with many people, this leads to slower processing, unnecessary detections, and extra manual cleanup—especially in large archives.

I propose adding user-controlled accuracy levels for face recognition, such as:

• An accuracy/clarity slider to set a minimum face quality threshold

• Options to detect:

  • Only clear and prominent faces

  • Moderately clear faces

  • Full detection (current behavior)

• Automatic exclusion of low-quality faces based on the selected level

Benefits:

• Faster processing in large archives

• Fewer false or low-value detections

• Better accuracy and user control

• More efficient workflows for institutional, journalistic, and humanitarian users

This enhancement would make Face Recognition more scalable and practical for professional, high-volume archives.