Facial Expression Recognition

The system labels each facial expression into 7 different expressions: neutral, joy, surprise, sadness, fear, anger and disgust.

All images are captured with a Logitech webcam. For each image, the eyes are first detected and the face boundaries are then estimated using a geometric face model. Afterwards, a set of facial features is extracted from the spatiotemporal domain and used to label each facial expression. The AdaBoost.MH algorithm is used here for multi-classification.

Category:

Science & Technology

Tags:

• Facial
• Expression
• Recognition
• LBP-TOP
• AdaBoost
• Machine
• Learning

[chinnikrishna55]

Where did you get the training data from

[ggiorgana]

@chinnikrishna55 I used the AU-Coded Facial Expression Database of Jefrrey Cohn & Tadeo Kanade.

[outfile]

very cool.

is this out-of-the-box or researching that you are doing?

[ggiorgana]

@outfile Thanks! This is researching that I did during my master in autonomous systems. The system is partially based on the approach proposed by Zhao and Pietikäinen in 2007 "Dynamic Texture Recognition Using Local Binary Patterns with an Application to Facial Expressions".

[auroradebonita]

@ggiorgana I am trying to find a program that will code facial expressions during research experiments. Is this program available? Do you utilize the Facial Action Coding System to derive your prototypical facial expressions?

[agungvibi]

I'm making a thesis about this, can you give a code for me?

thank you

agungvibibasqori@yahoo.co.id

[agungvibi]

can you send me your paper please.

into my email agungvibibasqori@yahoo.co.id

i'm making paper like you did.

tanks before :)

[Rocketboy5]

Hello, do you know where I can get source code for doing facial expression recognition in real-time? Thank you!

[HichamBadriTOBIS]

Well done