challenge-bot.com Git - ozzloy@gmail.com/oble/blob - TODO

   1 FOR SCREENSAVER:
   2 * use [Un]Inhibit together as a SimulateUserActivity workalike
   3         * disablegss.py from http://ubuntuforums.org/showthread.php?t=284804
   4         * this method doesn't work either.  tried in gss-poke.py
   5 * xdg-screensaver suspend / resume to SimulateUserActivity
   6         * http://tronche.com/gui/x/xlib-tutorial/2nd-program-anatomy.html
   7         * http://portland.freedesktop.org/xdg-utils-1.0beta3/xdg-screensaver.html
   8 * SimulateUserActivity
   9         * does it work on other people's machines?
  10                 * find if other people have their gnome-screensaver run even if they set
  11                 screensaver timeout to 1 minute and run:
  12                 for ((i = 0; i < 999; i++)); do
  13                         sleep 50;
  14                         gnome-screensaver-command --poke;
  15                 done;
  16         * find irc channel to ask if my understanding is correct
  17         * file bug
  18         * track down bug
  19                 * compile gnome-screensaver
  20 * cli opt parse
  21         * quiet mode
  22                 * no display window
  23                 * no text printed to terminal
  24 * fire up the face detector only just before going to inactive mode.
  25         * that way the camera isn't running all the time
  26         * maybe start 10 seconds before going inactive, then run until activity
  27 * be more aggressive with turning off the screen.  dim the screen as soon as
  28 it's not being looked at.  turn it off soon thereafter.
  29         * could be enough to save battery.  would need to power the camera sometimes,
  30         but would save on backlight.  need to test.
  31 * don't prevent sleep when face is detected, but user isn't actually looking at
  32 screen.
  33         * requires pose estimation.
  34         * also requires estimating the direction of the eye.
  35
  36 OTHER IDEAS:
  37 * authenticating a user based on face recognition (as opposed to face
  38 detection)
  39         * much more difficult to do reliably
  40         * use object tracking and emotion estimation to require a set of facial
  41         expressions as authentication.
  42                 * awesome for people who find typing hard, eg: no hands