Facial Recognition - customize your mirror for every user!

mc_landin

I am struggling to get modules using a name class to work. Still waiting for an answer as well

JavierTSalas

Hey guys. I’m having difficulty using my webcam with the capture.py script. I’m able to use fswebcame and motion to capture video but I keep getting errors running the script. I’m using a the playstion eye as a webcam. Any ideas?

salamures

Hi there, maybe someone is able to help me out.
I am trying to run capture.py but getting this Error

Traceback (most recent call last):
   File "capture.py", line 18, in 
      import lib.capture as capture
   File "/home/pi/MagicMirror/modules/MMM-Facial-Recognition-Tools/lib/capture.py", line 21, in 
     from . import config
   File "/home/pi/MagicMirror/modules/MMM-Facial-Recognition-Tools/lib/config.py", line 15, in 
     (CV_MAJOR_VER, CV_MINOR_VER, mv1, mv2) = cv2.__version__.split(".")
ValueError: not enough values to unpack (expected 4, got 3)

avinash

@salamures said in Facial Recognition - customize your mirror for every user!:

Hi there, maybe someone is able to help me out.
I am trying to run capture.py but getting this Error

Traceback (most recent call last):
   File "capture.py", line 18, in 
      import lib.capture as capture
   File "/home/pi/MagicMirror/modules/MMM-Facial-Recognition-Tools/lib/capture.py", line 21, in 
     from . import config
   File "/home/pi/MagicMirror/modules/MMM-Facial-Recognition-Tools/lib/config.py", line 15, in 
     (CV_MAJOR_VER, CV_MINOR_VER, mv1, mv2) = cv2.__version__.split(".")
ValueError: not enough values to unpack (expected 4, got 3)

I am also facing the same issue. Any effort to resolve it will be appreciated.

salamures

Are you using OpenCV 3.3? Then i might can help you.

salamures

Got another problem. My picam is not able to open. I am using opencv 3.3.

Remember to set the name list environment variable FACE_USERS
Using default FACE_ALGORITM: 1
Traceback (most recent call last):
File "facerecognition.py", line 30, in 
camera = config.get_camera(preview)
File "/home/pi/MagicMirror/modules/MMM-Facial-Recognition-Tools/lib/config.py", line 101, in get_camera
return webcam.OpenCVCapture(device_id=0)
File "/home/pi/MagicMirror/modules/MMM-Facial-Recognition-Tools/lib/webcam.py", line 29, in __init__
self._camera.open()
TypeError: Required argument 'device' (pos 1) not found

any suggestions?

PrivatVoidMarmot

Hello!

I have an problem wirth the Facial Regonition Module. I have tried it with the PiCam and an normal webcam. It worked not really good with die webcam and after i’ve tried it with die PiCam it doesn’t work any more. I’ve updatet the config.js.

This is the npm output:

config.js:

On the mirror nothing is shown. {
			module: 'MMM-Facial-Recognition',
			config: {
				// 1=LBPH | 2=Fisher | 3=Eigen
				recognitionAlgorithm: 1,
				// Threshold for the confidence of a recognized face before it's considered a
				// positive match.  Confidence values below this threshold will be considered
				// a positive match because the lower the confidence value, or distance, the
				// more confident the algorithm is that the face was correctly detected.
				lbphThreshold: 50,
				fisherThreshold: 250,
				eigenThreshold: 3000,
				// force the use of a usb webcam on raspberry pi (on other platforms this is always true automatically)
				useUSBCam: true,
				// Path to your training xml
				trainingFile: 'modules/MMM-Facial-Recognition-Tools/training.xml',
				// recognition intervall in seconds (smaller number = faster but CPU intens!)
				interval: 1,
				// Logout delay after last recognition so that a user does not get instantly logged out if he turns away from the mirror for a few seconds
				logoutDelay: 15,
				// Array with usernames (copy and paste from training script)
				users: ['Jenny'],
				//Module set used for strangers and if no user is detected
				defaultClass: "default",
				//Set of modules which should be shown for every user
				everyoneClass: "everyone",
				// Boolean to toggle welcomeMessage
				welcomeMessage: true
			}

Any suggestions?

thebeetleuk

@mc_landin I’m at this stage as well! for some reason the classes dont change whats displayed when a new face is seen. So in this example the “default” is seen all the time and then when a generic face is seen then the “everyone” class isnt shown. Has anyone got this working?

            {
                    module: "updatenotification",
                    position: "top_bar",
                    classes: 'default'
            },
            {
                    module: "clock",
                    position: "top_left",
                    classes: 'everyone'
            },
            {
                    module: 'MMM-Facial-Recognition',
                    config: {
                            recognitionAlgorithm: 1, // 1=LBPH | 2=Fisher | 3=Eigen
                            // Threshold for the confidence of a recognized face before it's considered a
                            // positive match.  Confidence values below this threshold will be considered
                            // a positive match because the lower the confidence value, or distance, the
                            // more confident the algorithm is that the face was correctly detected.
                            lbphThreshold: 50,
                            fisherThreshold: 250,
                            eigenThreshold: 3000,
                            useUSBCam: true, // force the use of a usb webcam on raspberry pi (on other platforms this is always true automatically)
                            trainingFile: 'modules/MMM-Facial-Recognition-Tools/training.xml', // Path to your training xml
                            interval: 2, // recognition intervall in seconds (smaller number = faster but CPU intens!)
                            logoutDelay: 15, // Logout delay after last recognition so that a user does not get instantly logged out if he turns away from the mirror f$
                            users: [], // Array with usernames (copy and paste from training script)
                            defaultClass: "default", //Module set used for strangers and if no user is detected
                            everyoneClass: "everyone", //Set of modules which should be shown for every user
                            welcomeMessage: true // Boolean to toggle welcomeMessage
                    }
            },

thebeetleuk

This post is deleted!

thebeetleuk

Hi All, I have found out that if you want to use your USB webcam for training you can update the file https://github.com/paviro/MMM-Facial-Recognition-Tools/blob/master/lib/config.py in the appropriate directory with the following. for some reason it doesnt seem to want to fall over to the webcam so I just commented out the pi web cam and added the usb one.

def get_camera(preview=True):
    try:
#        from . import picam
#        capture = picam.OpenCVCapture(preview)
#        capture.start()
#        return capture
        from . import webcam
        return webcam.OpenCVCapture(device_id=0)
        capture.start()
        return capture
    except Exception:
        from . import webcam
        return webcam.OpenCVCapture(device_id=0)

thebeetleuk

OK after lots and lots of trial and error I hope this helps people… these are the settings and results I had and as far as I can tell the everyone setting doesnt do what I would expect…

I don’t hink you have a way of identifying a stranger and do something specific for them?

defaultClass: "default", //Module set used for strangers and if no user is detected

It looks like when everyone is specified on its own then the logic doesnt seem to work. See classes: “everyone” below…

                                users: ['martin'], // Array with usernames (copy and paste from training script)
                                defaultClass: "default", //Module set used for strangers and if no user is detected
                                everyoneClass: "everyone", //Set of modules which should be shown for every user
                                welcomeMessage: false // Boolean to toggle welcomeMessage
                        }
                },
                {
                        module: "clock",
                        position: "top_left",
                        classes: "default" // displays with no face detection (OK) - not displayed when martin detected (OK) - displayed with no specific user detected (OK?)
                },
                {
                        module: "clock",
                        position: "top_right",
                        classes: "default everyone" // displays with no face detection (OK) - displayed with no specific user and martin (OK)
                },
                {
                        module: "clock",
                        position: "bottom_right",
                        classes: "default martin" // displays with no face detection (OK) - displayed with no specific user and martin (OK)
                },
                {
                        module: "clock",
                        position: "bottom_left",
                        classes: "everyone" // never displays (??? I would expect this to display when Martin is detected - or any other user)
                },
                {
                        module: "compliments",
                        position: "lower_third",
                        classes: "martin" // only displayed when martin detected (OK)
                },

gismo2006

Hi,

I use your Facial Recognition modul. I have create some user and everything work fine.

But now I install MMM-Motion-Detection modul. I works fine, too. But if I use this modul your modul didn´t work. When I cut off the MMM-Motion-Detection module from config.js and start again your modul will work fine again.

Can someone help me?

greets gismo

PrivatVoidMarmot

@paviro I have the same problem like @shashank, but I don’t figure out any mistake… the trainingfile is in the right direction… It doesn’t work with PiCam and with an webcam neither…

PrivatVoidMarmot

Sorry for spaming,

but I tried to refresh everything. So I installed the Rasbian new, the Magic Mirror, the Facial Recognition and the Tools. It worked… ONCE. After rebooting the system I have the same error as bevor:

It load training data and select the PiCam and then it doesn’t go on. No error-message.

Any idea which component was active when I reinstalled everything and isn’t active after rebooting?
The camera is enabled.

jer78

I have the same issue. Training went fine. I got everything working but then after a reboot it seems to load the module but then nothing. It doesn’t seem to want to do anything. No errors just no facial recognition.

jer78

So I figured out my own problem. Seems like it doesn’t like it when I changed the language file. Restored the old one and everything works as it should. Also, I find raising the threshold to 250 for the LBPH Algorithm helps a lot to get more accurate results. The confidence seems to hover around 49-52 for me but I couldn’t get any results with 50, constantly thought I was “stranger”. Thanks for all the hard work on this module. It’s really fun to play with!

Prestitivo

I got this module working after a bit of trial and error, and it’s awesome. I only have 1 issue so far. When I’m recognized, all of the modules drop by about 30 pixels. I’m not a programmer by any stretch of the imagination, but have no issues looking around at code trying to figure out what it does. I used Firebug to see what was going, and from what I can see, the one of the div’s updates from style=“display: none;” to style=“display: block;”. The part that’s really confusing me is that it only does this when it recognizes me. When it recognizes a stranger, none of the modules shift. I can provide anything requested, but I wasn’t sure if anyone else was experiencing this issue.

j.e.f.f

@Prestitivo It means you’ve configured that module to be displayed when you are recognized, and hidden when you are not. Take a look at your config to see if you can pinpoint which module is making this happen.

Prestitivo

Ah yes, thank you. It was the “updatenotification” module in the top bar. I took myself off of that, and now it works as expected. Now that I think about it, that makes sense.

Still have a lot of tinkering to do, but this solved the annoying issue I caused.

Thank you kindly for the help.

WookyDO

Hi… will there be any possibility to use Facial Recognition along with motion detection for turning on/off the Display ?

Regards,
Wolfgang