Personal tools
You are here: Home Scientific Research Projects Image Spam Classification

Image Spam Classification

An Email anti-spam tool consists of a succession of filters through which a message passes. Each filter focuses on specific criteria to determine whether the message is spam or not. Today, spam designers manage in some cases to circumvent these filters. Fastnet is a world leader in the market of anti-spam.

The objective of the project is to provide MailCleaner, Fastnet's flagship product, with additional filters based on image recognition. They will use image processing algorithms such as classification by signature extraction and distance computation. This new approach will enhance the set of filters, currently used by MailCleaner products and give Fastnet a new technology advantage against its competition. 

Partners

Document Actions
Project Information
Themes: Perceptual and cognitive systems
Funding: Commission for Technology and Innovation (CTI)
Shortname: IMPACT
Dates: Start date: Jan 01, 2009
End date: Mar 31, 2010
Contact: Jean-Marc ODOBEZ
+41 277 217 726