This is a really interesting topic to me. I was briefly introduced to some machine learning research that seeks to generate 3d models of objects given very few (often only a single) images. It's intriguing to think about the types of algorithms/techniques that can allow models to learn such a complex. For anyone interested, one such model is PIFuHD, and here's the paper/code
Staffyirenng
Thanks @wangjim19! Thanks @Sicheng-Pan! I edited your post slightly to use Markdown to create the hyperlink. We encourage students to use Markdown on our website for that, as well as to write equations. We have a 90 second primer on getting started -- it's easy!
atsu-kotani
I immediately connected this with Animoji by Apple few years ago (e.g. https://youtu.be/UMiheAU-ABs).
This is a really interesting topic to me. I was briefly introduced to some machine learning research that seeks to generate 3d models of objects given very few (often only a single) images. It's intriguing to think about the types of algorithms/techniques that can allow models to learn such a complex. For anyone interested, one such model is PIFuHD, and here's the paper/code
Thanks @wangjim19! Thanks @Sicheng-Pan! I edited your post slightly to use Markdown to create the hyperlink. We encourage students to use Markdown on our website for that, as well as to write equations. We have a 90 second primer on getting started -- it's easy!
I immediately connected this with Animoji by Apple few years ago (e.g. https://youtu.be/UMiheAU-ABs).