Head Reconstruction from Internet Photos

TL;DR

This paper introduces a method for reconstructing the rough shape of entire heads from uncalibrated Internet photos by gradually growing a head mesh using photometric stereo constraints, extending beyond just the face.

Contribution

It presents the boundary-value growing algorithm that reconstructs full head shapes from Internet photos, addressing a gap in existing face-focused methods.

Findings

Successfully reconstructs head shapes from celebrity photos.

Extends head modeling beyond face to include full head shape.

Demonstrates effectiveness on uncalibrated Internet images.

Abstract

3D face reconstruction from Internet photos has recently produced exciting results. A person's face, e.g., Tom Hanks, can be modeled and animated in 3D from a completely uncalibrated photo collection. Most methods, however, focus solely on face area and mask out the rest of the head. This paper proposes that head modeling from the Internet is a problem we can solve. We target reconstruction of the rough shape of the head. Our method is to gradually "grow" the head mesh starting from the frontal face and extending to the rest of views using photometric stereo constraints. We call our method boundary-value growing algorithm. Results on photos of celebrities downloaded from the Internet are presented.