Facial Recognition.

Imagine, you have thousands of family photos on old compact discs and you want to sort them by date and facial recognition and then arrange into new albums. You don't like to upload your photos to online resources from a privacy perspective or another reasons. The updated Find option of CDRoller allows to search and sort your photos by facial recognition.

Key Features.

  • Detects multiple faces in an image.
  • Calculates the matching threshold at a given False Acceptance Rate (FAR).
  • Examines each photo on the selected drive (folder), matches the face template, acquires and displays the facial similarity level in %.
  • Allows to rotate the examined image for a better face detection.
  • Sorts the found photos by taken date.
  • Supports digital camera RAW formats, such as: Canon (*.cr2, *.crw), Kodak (*.dcr), Minolta (*.mrw), Nikon (*.nef), Olympus (*.orf), Pentax (*.pef), Fuji (*.raf), Leica (*.raw), Sony (*.srf) and Sigma (*.x3f).
  • Works under Windows Vista/7/8/8.1/10/11 operating system.

Face detection and recognition is integrated into the Find option of CDRoller. This allows to search the entire disc or specific folder(s) for the face template(s). To activate the Search window, just press F3 button or use the Find command from the Tools menu, or select multiple folders and click the Find in Selected in the Other Tools section of Side Bar (a left panel of the program window). Then, go to the Face tab.

Click the Image for template(s) button and assign a picture where the face template(s) will be extracted from. We suppose you already have some pictures (candidates for that) in your computer.

Face recognition. Select face template

 

Click the Start button, wait for the end of search and look at the results. If no faces found, try to change some settings under the Face tab. For example, try to decrease Minimal Face Quality or set Internal Resize Width to 512 for the bad quality photos.

Face recognition. View search results

 

To view the found pictures, select the displayed file(s), and use the View Image command.