Open shanealv opened 4 years ago
Update:
If the stop condition is set to PAUSED
instead of STOPPED
, it does work. At least, a face will be removed. So the state is changing, just not in a way the example code expects. And of course, we still are only getting one face tracked.
SPECIFIC ISSUE ENCOUNTERED
Using the following code for getting a list of augmented faced from the camera:
I am consistently only getting 0 or 1 face detected, as output logs have indicated:
This is despite having two faces in frame.
In fact, if I cover one face, tracking immediately swaps over to the second face, and the initial face is subsequently ignored until the second face is covered.
As per the sample code, I have a section for adding faces as they appear ready for tracking:
It appears that the conditional is only satisfied once, resulting in the following log message appearing once:
I also have this as the code to stop tracking:
Which doesn't print to the log at all, even when I've clearly put my phone down and I get repeated "detected: 0" log messages.
TL;DR: Getting all trackable faces is really only giving my one, and that face is seeming always in a
TRACKING
state.VERSIONS USED
1.14.191118106
On Windows, use:adb shell pm dump com.google.ar.core | findstr /i "packages: versionName"
On macOS, use:adb shell pm dump com.google.ar.core | egrep -i versionName\|packages:
adb shell getprop ro.build.fingerprint
: `OnePlus/OnePlus6/OnePlus6:10/QKQ1.190716.003/1912142009:user/release-keysSTEPS TO REPRODUCE THE ISSUE
WORKAROUNDS (IF ANY)
n/a
ADDITIONAL COMMENTS
Is there perhaps some documentation I've missed marking this as a limitation? The examples and API suggest that multiple faces were meant to be detected, although clearly that isn't the actual behavior in this case.