Andreas Lang, Marc Ritter (2012). Face Detection using Swarm Intelligence (Gesichtsdetektion mittels Schwarmintelligenz). In: Junge Wissenschaft doi:10.17605/OSF.IO/XG66G.
As final project of my high school Kepler Gymnasium Chemnitz I worked on Face Detection using Swarm Intelligence, supervised by Marc Ritter at the Chair Media Informatics. I won the first prize in the area mathematics and computer science of the German national contest for young scientists Jugend forscht (more than 10 000 participants per year) as well as the “Europapreis” of the German Research Foundation DFG. Furthermore I was delegated to the 22nd European Union Contest For Young Scientists where I won the special prize “Vision Box”. Additionally, I presented my work at an exhibition in the Hessian representation at the EU in Brussels, and at the Expo Sciences of Jonk Fuerscher Lëtzebuerg in 2011 an 2012.
Summary: Face recognition in a computationally expensive problem in image processing. This process can be accelerated by inputting position and size of a face in an image. The goal of this project is to create an efficient algorithm for such face detection. The proposed algorithm initially distributes particles on the image. They target a certain range of hues and their movement is directed by the concept of swarm intelligence, i.e. behaviors such as opportunism, avoidance and memory. The algorithm is extended to moving images by taking into account the particle positions of the previous image, thus saving computation cost. For visualization, a 3D hat is added on top of the detected face. The proposed algorithm serves as an accelerator for face recognition algorithms.
The paper is publicly available at http://www.qucosa.de/fileadmin/data/qucosa/documents/6441/EUCYS%20-%20Face%20Detection%20using%20Swarm%20Intelligence_Final.pdf.