It can be useful for research if we make this software available on the web and produce a database of stored data points. If we ask users to supply their age, gender, nationality, and race, we can compile averages of the data points as more data is gathered. This can potentially be used to answer certain visual queries such as “How do facial feature ratios differ with age group, gender, or race?”

With the help of some image processing algorithms, it would be useful if users did not have to supply data points themselves. If the algorithm can automatically provide the points, then there will be no chance of human error.

With the correct groups of images, we can extend our visualizations to apply to 3D models of faces instead of just front-facing images.

After our initial design we received good constructive feedback related to the aspects of our visualizations that were effective and those that required some work. We reviewed all of our designs and produced new and improved visualizations that took into account the downfalls of our old ones. In our second presentation we had good feedback to the changes we made, with a few last minute suggestions to improve our designs even further. The changes we made after this were:

Make the comparison visualization use two distinct transparent textures it improve the representation of areas of difference

The line widths of the symmetry visualizations were changed to accentuate differences

A legend was added to the symmetry visualization to map the data to it’s corresponding position on the face.

Our final visualizations address all of the visual queries that we identified in our initial scope. They are simple and easy to read, and together they represent all of the data collected in the user input phase.