Introducing CullAi - Ai assistant for professional photographers to select final photos from a photoshoot.

While selecting best photos from a photoshoot is subjective and will require human touch, CullAi assists in this process by filtering objectively bad quality photos and selecting relatively better quality ones from group of similar photos.

CullAi currently focuses on majority professional photoshoots which have people as their primary subjects. It prominently uses Face Capture Quality which is index of better lit, sharper face with open eyes.


CullAi filters photos with objectively bad quality including:

  • Bad lighting conditions
  • Out of focus primary subjects
  • Camera shake
  • Closed eyes

Manually checking this for every photo is time-consuming.

CullAi Filters


You can adjust Face Capture Quality threshold to filter out photos with lower overall quality. Threshold is applied to the face having the highest Face Capture Quality among all faces in that photo.


CullAi groups similar photos with:

  • Similar location
  • Similar background
  • Same subjects

Grouping of similar photos enables you to compare them side by side or with blending without any manual effort.


You can adjust Similarity Sensitivity and Similarity Time Gap to influence this. Higher sensitivity will group more photos together which might be less similar.


CullAi by default selects photo with the best overall face capture quality of people in that group of photos. As this is a subjective choice you can select a different photo manually also.


Many times in events like weddings, photos will have many people besides primary subjects of the shoot.

CullAi automatically treats people who are part of more photos as more important, but user can change this using thresholds given for each detected person. This is required so that one out of focus person in background shouldn't influence better photo in a group.


CullAi Demo Video


CullAi is available for macOS for now, we will release same for other platforms based on feedback.