| Figure 3: Details of step 2. (a) Input image and (b) intrinsic contours and parts of the texture, here represented by . Note in that although there are texture losses close to the neck and background, face intrinsic contours are preserved, such as eyes, mouth, and hair. |
|