Profile cover photo
Profile photo
Blake Senftner
2,569 followers -
Technology Entrepreneur, MBA, Web Ninja, Mad Scientist
Technology Entrepreneur, MBA, Web Ninja, Mad Scientist

2,569 followers
About
Blake's posts

Post has attachment
Complement to my just posted YouTube video:

Post has attachment

Post has attachment
Available now: full body, instantly created from one photo, Fuse 1.3 compatible 3D Avatars ready for use in Maya, Unity, Unreal and more. Once exported and in Fuse desktop, wigs, clothing, more body deformations, rigs, and motions. www.3D-Avatar-Store.com
Photo

Post has attachment
As usual, been working on the reconstruction system. Getting ready to release new full body Fuse 1.3 compatible 3D Avatars, each generated from a single photo.
Photo

Post has attachment
Just released the first version of our Developer Console at the 3D Avatar Store. With the Developer Console, one can view, run, edit and re-run HTML / CSS / Javascript source code examples, as well as create completely new, private and secure apps of your own using the 3D Avatar Store's WebAPI. Integrate neural net powered 3D Reconstruction of facial photos into auto-magically lip sync'ing 3D Avatars looking just like the person in the supplied photo.

Check it out! Create talking avatars of your family and friends, put words in their mouths, and more. There's no better platform for learning 3D animation development than when applied to a little bit of mischief!

Post has attachment
Been working hard at the 3D Avatar Store, so many new features I forget writing them all. Here's a post describing our fantastic new Facial Feature Finder, showing some of it's great results, plus the first peek at our new Developer Console.

Post has attachment
I've been working in a new neural net generation, updating from generation 2 to generate 5 at www.3d-avatar-store.com. These are the 'nets that perform facial feature identification in photos, and the difference between the generations is remarkable. Where the previous facial feature identifier has a training set of 70,000 real people's photos and laser scans, the new generation has 300,000 subjects, the facial feature identification quality is significantly improved,  the processing time is 5% what it was previously, and this generation works with video - able to analyze multiple video frames and generate both a "best guess" at a person's true 3D shape as well as the individual's frame-to-frame expression distortions as portrayed in the video source. I'm very excited to bring this new capability online.

Post has attachment
Recent update to my automated 3D avatar creation platform: Auto 3D Avatar Creation Demo

Post has attachment
Been working on this for a while: given a photo of someone it performs 3D Reconstruction and generates an animation ready photo-real avatar. Just added auto-generation of 82 facial morph targets, and a small library of auto-shrink-wrapping wigs, ears and other accessories for in-browser play. This is a Web App as well as an API, so you can include Avatars as well as have performance ready 3D Avatar creation in your own apps and games.
Wait while more posts are being loaded