Hello, face recognition is already part of ARC.

Navigation can be done by combining GPS and Compass add-ons. Both of which already exist also.

:)

Face recognition i mean without gyphs. The camera take screen capture my face and if camera meet me then report my name. There is another project or video with navigation map? Thanks.

You can download ARC by selecting that link of pressing Downloads from the top menu. Load ARC and perform the following:

a) Select Add Control b) Select Camera c) Select Camera Device d) from the Devices list, select your video capture device e) Select the Tracking tab f) Select Face Tracking g) You may now place your face in front of the camera and it will be tracked h) There are tutorials on how to use EZ-Script within the Config menu for Start and Stop tracking scripts

As for navigation, the EZ-Script commands are pretty straight forward. One day we will do a project video, but it's not a high priority because GPS is not accurate enough for a home or yard, unless your home is incredibly massive with no roof .. The navigation with GPS and Compass is for outdoor robot challenges and navigation through terrain - it's not generally used for home use.

To understand what room your robot is in, you can get creative and use Glyphs for each room. However, we'd one day like to have a module to allow your robot to understand what room it is in - but that's a ways away for today's technology. Any swarm/navigation videos you see on youtube use internal RF/IR beacons, undercarpet signals, and other methods.

My suggestion is to follow the Tutorial section and learn the basics of ARC before jumping into navigation and such.

@Daedalus Hello and welcome. ARC/SDK has face tracking already, however it will not "recognize" a particular face... yet? I'm just barely scratching the surface of understanding with this HAAR stuff.

Currently, the closest you might get for a recognition response is via the glyph feature, but utilizing the glyph as part of an ID badge or something to go along with the face tracking. See a face? = YES, see a glyph? = YES, if both answers YES then match glyph to pre-programmed response. Not true identification, but better then just swiveling the camera around