~/talk

Augmenting Faces and Images

on YouTube 35m
Talkat Google I/O 2019 · 2019-05-07..05-09· [webxr, computer-vision]

speakers

name:Ashish Shah
links:none on file
name:Anna Miyajima
links:none on file
name:Ranjith Kagathi Ananda
links:none on file

resources

video:https://youtu.be/TGtFqe4s51o
slides:none on file

summary

ARCore ships concrete APIs: a 468-point face mesh plus center and region poses that run on Android and iOS without depth hardware, powered by TensorFlow Lite transfer learning, and claimed to work on over one billion devices. Augmented Images supports up to 1,000 reference images, now returns 30 Hz image poses (15× more frequent) with under 3% extra power, and shows real demos (Meitu filters, JD AR Books, Notable Women) that track, attach, and animate virtual content reliably.

AI-written summary · about

chapters (4)

  1. 0:00Start
  2. 5:32Anna Miyajima Google
  3. 5:54Ranjith Kagathi Ananda Google
  4. 18:24Ashish Shah Google

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.