Narcissus

MLP-based facial expression recognition tool

Description

Narcissus is an application that uses machine learning techniques (artificial neural network with 2 layers of weights) to classify static frontal facial images into Ekman's 6 emotion categories with clear facial signals (anger, disgust, fear, happiness, sadness, surprise).

Download

March 2, 2005

Narcissus 0.9_RC2 has been uploaded. Addition of modules is now done automatically; a huge improvement. This is not reflected in the graphs, though, and maybe it never will be because I think it's clearer this way.

If you want to try it out, download one of the packages below. Make sure you have Java 1.4.2 or newer in your path (older versions might work too, I haven't really tested). Unpack the archive and type

	java -jar Narcissus.jar	
	

optionally followed by some arguments, and off you go!

The following packages are available (source code and Liquid theme included in all packages):

You can find some written material about Narcissus in the publications section.

W3CCSS
W3CXHTML 1.1