Graphics and Media Lab

Motivation

Face detection and tracking has been the topics of an extensive research for the several past decades. Many heuristic and patternrecognition based strategies have been proposed for achieving robust and accurate solution. Among feature-based face detection methods, the ones using skin color as a detection cue, have gained strong popularity. Color allows fast processing and is highly robust to geometric variations of the face pattern. Also, the experience suggests that human skin has a characteristic color, which is easily recognized by humans. So trying to employ skin color modelling for face detection was an idea suggested both by task properties and common sense.

Color image and detected skin

Numerous techniques for skin color modelling and recognition have been proposed during several past years. A few papers comparing different approaches have been published. However, back in 2003 a comprehensive survey on the topic was still missing. In 2003-2004 a project for studying skin detection methods was done.

The survey

Our goal in the first phase of this project, was to gather published techniques, describe their key ideas and try to find out and summarize their advantages, disadvantages and characteristic features.
The results of the first phase are:

1. A survey:

   Vezhnevets V., Sazonov V., Andreeva A., "A Survey on Pixel-Based Skin Color Detection Techniques". Proc. Graphicon-2003, pp. 85-92, Moscow, Russia, September 2003.
   .pdf (93 KB) .ps.gz (63 KB)

   Included in CV-Online: On-Line Compendium of Computer Vision. (Remote copy) ,
   Cited in 68 papers according to "Google scholar";

2. An almost complete collection of papers on skin detection:

   Skin detection papers (the library interface is in russian, but very easy)

Numerical evaluation

The second phase of the project, was devoted to numerical evaluation of different methods on common dataset. The result was a huge technical report, that we publish here (a draft version).

Skin cluster in nRG space (top view)	Skin cluster in nRG space (side view)

Team

• Vladimir Vezhnevets, Lomonosov Moscow State University
• Vassiliy Sazonov
• Alla Andreeva