They're releasing a patch for that soon for headphone recognition.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Real-Time Motion Capture Firm Faceshift May Have Been Acquired by Apple
- Thread starter MacRumors
- Start date
- Sort by reaction score
It seems entirely possible that Apple are making their own VR or AR headset. With Microsoft Hololens, Facebook's Oculus Rift and Google's ( part funded ) Magic leap.
Real time face mapping has a lot of applications for games and other apps.
Real time face mapping has a lot of applications for games and other apps.
Shake was fantastic at the time. But it was very far behind the curve in a short space of time. Fusion and Nuke had it beat and a more open architecture and standard plugins. And the main problem was most of the key guys went to ILM and Pixar and was in dire need of a complete re-write.
FaceTime: Hologram to hologram. Now you can be completely naked or doing unspeakable things, and always look your best! [insert hyperbolic epilogue here, using the word "Magical" at least once]
Am I the only one who immediately thought of this when I heard about real time motion capturing?
https://www.macrumors.com/2015/04/14/apple-acquires-linx-imaging/
They acquire a company that makes a tiny cheap camera capable of 3d capture, and then a face recognition company. I'm not entirely sure what they want to do with that (there are plenty of possibilities) but it definitely seems related to me.
https://www.macrumors.com/2015/04/14/apple-acquires-linx-imaging/
They acquire a company that makes a tiny cheap camera capable of 3d capture, and then a face recognition company. I'm not entirely sure what they want to do with that (there are plenty of possibilities) but it definitely seems related to me.
I'm sure there will be other info (like 3-dimensionality) gleaned by the camera to prevent that.
It's make an interesting thing to keep in the closet, especially when company comes over.
Simplest things would be FaceID and Head/Eye tracking for interaction.
More complex would be "emotion" reading from facial expressions. Depending on how robust the capture is we may start seeing devices that can pick up on very subtle changes in attitude. To me there are possibilities in education. A big chunk of a teacher's skill set comes in with being able to read student's attitudes. It's something at currently no automated system can do, but if it could it could mean major things for self-directed/computer-driven learning devices. Not that I see Apple being the ones to do it, very unlikely there, they tend to be rather blind to such things in their "single consumer hyper focus" madness of the last few years.
That's not so much a problem of the face recognition software but rather a problem of the 3d models.
Also, I don't think the idea is even supposed to be to translate the actors face as close possible into the CG face. The actors have real time feedback and that's the achievement. I imagine it like those drawing tablets. At first it's really weird, because you draw something on a tablet and something happens on the screen right in front of you. But after a while your brain "gets it" and can compensate, so the movement of your hand is in sync with what happens on the screen. The same will happen for the actors. They look at the CG faces and not at their actual faces and quickly learn how their facial expression translates and can compensate. So it's useless to even compare the real and CG image, because the actors know what the end result will look like. It's actually the same (if not worse) when actors wear extensive make-up. They can't just act as they would without any make-up on. They need to compensate for how the make-up distorts their facial expressions.
Seems like there's a lot of scope for User Interface with this. Facial gestures like smiling, swallowing, biting, nodding, head shaking, blinking or winking could trigger events. Eye movements could move documents and images around. If it could work out which part of the document you are looking at, it could automatically scroll when you get to the bottom of the page. By reading your mood, it could generate music to match the way you're feeling. It could highlight which parts of an article made you laugh, or surprised you, or interested you.
The possibilities are endless.
The possibilities are endless.
could be the problems of models though. Wish they would show just mesh and tracking points.
You must of watched something different to me as that was an awesome demo. Face recognition and motion tracking just a few years ago aas a massive labour for very poor results.
The first part was a little meh but the superhero motion test section was very good and showed a vast array of facial expressions. The computing that takes to do in real time with that precision is still staggering and it handled it very well.
Apple has been looking more in to its own media creation of late and games are the biggest seller on the app store so it may well be Apple will start producing their own. Or working with select developers.
Nobody else has, so I'll throw in the obligatory:
Sesame released this earlier this year. Target audience so far...paralyzed individuals.
Or maybe create an API / SOC co-processor and bake it in the ecosystem. That way it increases the ecosystem's stickyness and helps devs make money.