We don't do face detection in real-time, it's done for cropping photos. I'm not aware of any profiling that's been done on performance.
The memory copy isn't really a binder issue, it's due to the way Java handles byte arrays. On Dec 29, 11:52 pm, "formos...@gmail.com" <formos...@gmail.com> wrote: > Hi, > > Current face detection is called from java code, is that possible to > call it from mediaserver? > For example, in camera service, if we can detect face in preview > callback, we can display the recognized face rectangle on preview > frame and avoid extra memory copy between mediasever and java > activity? > > From my understanding, preview frame is a pmem and can be passed to > another processes by ibinder, there will be memory copy when calling > face detection in java, right? > > I'm also curious the performance of neven library in android, can > anyone share experience? > > Thanks. > > Formosalf --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "android-framework" group. To post to this group, send email to android-framework@googlegroups.com To unsubscribe from this group, send email to android-framework+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/android-framework?hl=en -~----------~----~----~----~------~----~------~--~---
