I spent years developing and then offering a system that remotely enabled creating realistic 3D avatars of people from a single photo. It works, I scaled it to viral capacity levels, the quality is high (https://twitter.com/3davatarstore), and my prices were nearly free. Yet, game, VR and VFX studios only wanted it free. After being jerked around for years by circuses of clowns, I shut it down. I'm happier doing FR now for government agencies. VR will continue to be a lonely place until the corporate exploitation is regulated to the degree that an individual will have legal and portable ownership over their appearance in VR technologies, with all the legal ramifications that exposes.
Interesting! Are you able to share anything about how your algorithm works? I've been playing around with feature detection (using HoG descriptors) plus constrained local models to identify locations in an image and then manipulating the mesh to better match these locations and extracting a corresponding texture.
I dunno if it's fair to blame corporate exploitation for your experience - I feel like the current perception that everything must be "free - maybe with ads, and maybe pay later" is more to blame. :/
Portable personal avatars is going to be a huge undertaking. I'm really looking forward to the results, though!
