Methods for forming composite face images

A method for forming a composite face image with a processor executing computer executable instructions, the method comprising the steps of:

• receiving into a memory unit a first image file including a first face image;
• receiving into the memory unit a second image file including a second face image;
• globally aligning orientations and sizes of the first face image and the second face image;
• locally aligning the first face image with the second face image to vertically align facial features of the first face image with facial features of the second face image;
• creating a composite image file having a virtual canvas for depicting the composite face image;
• inserting a left half of the first face image on a left side of the virtual canvas;
• and
• inserting a right half of the second face image on a right side of the virtual canvas adjacent to and aligned with the left half of the first face image about a seam between the left half of the first face image and the right half of the second face image;
• wherein locally aligning the first face image with the second face image includes:
• selecting a target feature of one of the first face image and the second face image, wherein the target feature is either a pose or a shape of a face;
• and
• transforming the first face image or the second face image with a generative adversarial technique to match the target feature.

A method for forming a composite face image with a processor executing computer executable instructions, the method comprising the steps of:

• receiving into a memory unit a first image file including a first face image;
• receiving into the memory unit a second image file including a second face image;
• globally aligning orientations and sizes of the first face image and the second face image;
• locally aligning the first face image with the second face image to vertically align facial features of the first face image with facial features of the second face image;
• creating a composite image file having a virtual canvas for depicting the composite face image;
• inserting a left half of the first face image on a left side of the virtual canvas;
• inserting a right half of the second face image on a right side of the virtual canvas adjacent to and aligned with the left half of the first face image about a seam between the left half of the first face image and the right half of the second face image;
• receiving into the memory unit an input image file;
• detecting the location of face images in the input image file;
• selecting a selected face image from the face images detected in the input image file;
• cropping the input image file to include only the selected face image;
• sending the cropped input image file to the memory unit to be received as the first image file;
• detecting the locations of selected facial features in the selected face image;
• determining selected geometric parameters of the selected face image based on the detected locations of selected facial features;
• saving a geometric parameters file including the determined geometric parameters into the memory unit;
• and
• associating the geometric parameters file with the first image file received in the memory unit.