EDIT 5/11/2017 - Video explanations ( inmoov use )

 

EDIT 4/11/2017 - Added unicode support

Added unicode support to this filter because opencv dont like it ( cannot use unicode inside folder names ).

Nothing will change about use, but want to explain the workarround :
Because opencv create a folder based on the trained name we need to map the trained unicode name to ascii name. I use an extra file called "THENAME".label , and use a random string for folder name. This is mapped inside internal hashMap

---------------

First report  :

Inside this test there are 4 faces. 3 of those faces are correctly parsed by cascade classifiers.

2 faces was trained ( Ryan + Rachel )

 

 

 

 

Script : https://github.com/MyRobotLab/pyrobotlab/blob/develop/home/moz4r/Test/F…

[[home/moz4r/Test/FaceRecognizer.py]]

2 problems :

1 clear face is not parsed ( nok-test2.jpg ).

1 face ( OK-test1.jpg is reconized as Ryan a previous trainded face. This guy is not ryan ( tested with other faces too, it seem something is forced when the classifier detect a no trained face.

kwatters

6 years 6 months ago

Moz4r, fantastic video!  You've got it basically all working.  There are some keys to how the current face recognition works.

1. as you know, you need at least 2 faces trained before it will recognize anyone.

2. it will always guess the face based on who it knows.  It does not have a concept of an unknown person.

3. face recognition only occurs when both the eyes , mouth , and a face is detected.  This is pretty strict, we could relax this.

So, it's not that face recognition is broken, it's just that , I think , we want it to work better :)

The original implementation that we did (with scruffybob's help) is very strict and doesn't attempt to make a face recognition prediction until it's very very confident that it's seen a face.  This is a decision that favors "precision" over "recall".  

the idea is that it will only predict a face being recognized in a very conservative manor, we can relax this requirement and in a way, I think that's probably a good idea....  maybe we could require a very strict match for training, and a looser match for recognition... 

That's one issue that you're probably seeing, the second issue is there is no concept of an unknown face with the current face recognition algorithm and I think that problem needs a bit more thought...

 

 

 

moz4r

6 years 6 months ago

Hi Kevin can you help me to implement concept of an unknown person I don't know where to start..

Goal is to use publishNoRecognizedFace(), store the face as temp, and ask the name.

I didnt find how the filter chose a random picture for an unknown person and fake the result