Download

Do you have any plans to use Generative Adversarial Networks to generate Meta Human content in the future?

The reason I am asking is that with this current tool-set, everyone will end up with similar looking characters in their projects. One way that the tool could be improved is that Instead of using predefined scanned data for skin textures, some of this data could instead be generated using GAN techniques to create unique content. The current level of customization appears to be rather very limited and even if more scanned data is added, the problem of creating unique content will remain. You can already see that most of the characters rely on the same limited data set. A more procedural approach, is probably the answer in my opinion.

https://thispersondoesnotexist.com/

Right now, the GAN technology is limited to 2D content, hence why it might make the most sense to use it for skin texture generation, but in the not so distant future, these concepts will be applied to 3D data as well. Meta Human combined with the power of GAN, would truly be a game changer. Over time, the same principles could potentially be applied to body types, facial features, facial hair and hair styles, and more of course (e.g non-humanoid creatures). But as a first step, using GAN techniques to generate unique skin textures would increase the probability that the characters that are created with the tool end up having a more unique look and feel (instead of relying on scanned data that only allows for a predetermined set of possible permutations).

Is the team behind this tool considering this approach? And if so, what can you tell us more about it and where the tool will be evolving in the future?

Just wanted to share my thoughts with the team and I am sure that what I have mentioned is already part of the discussions that you are having.

Thank you.

5 Likes