Webcam facial recognition

Question

I am currently working on a project where I need to extract the facial expression of a user (only one user at a time from a webcam) as sad or happy.

There are many face detection programs / APIs, but I have not found anyone who can recognize automatic recognition.

The best opportunity I've found so far:

I used to use OpenCV to detect faces, which worked just fine, so if anyone has any tips on how to do this with OpenCV, that would be great!

Some users in the OpenCV panel suggested looking for AAM (active apereance models) and ASM (active figure models), but all I found was documents.

-So I use models with an active form with Stasm , which will give me access to 77 different points in the face, so I still have to manually match their expressions.

We welcome any programming language.

Answer 1

After the release of 2.4, opencv appeared with the face recognition API. You can define sad and happy faces as two different faces and use the API to classify them. If you are looking for any tutorials, you can check the opencv tutorial page .

Answer 2

Perhaps this article may help you.

This is an explanation of the concepts used to achieve the result shown here :

In addition, you can look here .

All of the above is the work of people who more or less have achieved what you saw.

I want to do the same thing that you do - face recognition using OpenCV, so I hope that sooner or later we can exchange our ideas.

Answer 3

Half of the problem is solved for you using OpenCV. If you created OpenCV with examples, you can run an example called c-example-smiledetect , which can detect a smiling face with different levels.

The code should be in the OpenCV folder in samples / c . I remember you can adjust the sensitivity in the code so that this can help. You can also find the code here .

In another note, the implementation of AAM using OpenCV can be found in this book (chapter 7). This does not include part of the search and fit, but shows how to create AAM from ASM and texture.