Fraunhofer-Institut für Graphische Datenverarbeitung
For our Competence Center Smart Living and Biometric Technologies we’re looking for a
Biometrics is a rapidly growing technology that aims to identify or verify people identities based on their physical or behavioral properties. Face recognition is becoming one of the most used biometrics as the social acceptance met the recent advances in face recognition performance especially with approaches based on deep learning networks.
Biometric presentation attacks is one of the main barriers preventing the use of face recognition in high security unattended applications. However, this is usually studied under the assumption aim is to be recognized as a specific user (the attacked profile). This thesis tries to evaluate and propose solutions to spoofing blacklist system with the attackers aiming at hiding their identity.Was Sie mitbringen
- Interest in computer vision and machine learning
Informatics, electrical engineering, mathematics, physics.
Was Sie erwarten können
This thesis proposal is split into two continuous parts, the first part aims at evaluating state of the art face recognition algorithm under anti-identification presentation attacks. This include designing an evaluation process, design and develop a testing platform, collecting data, evaluating a number of standard face recognition approaches, present and discuss the achieved results. The second part Selects a subset of the evaluated approaches and based on analyzing the previous results, tries to modify these approaches to be more robust against these attacks without losing much performance. The proposed solution should be evaluated and the achieved results should be thoroughly discussed. For a bachelor thesis, only the first part is included. A master thesis should contain both parts.Key literature:
 B. Amos, B. Ludwiczuk, M. Satyanarayanan, "Openface: A general-purpose face recognition library with mobile applications," CMU-CS-16-118, CMU School of Computer Science, Tech. Rep., 2016.
 Naser Damer, Kristiyan Dimitrov: Practical View on Face Presentation Attack Detection. BMVC 2016.
 Ivana Chingovska, J. Yang, Zhen Lei, Dong Yi, Stan Z. Li, Olga Kähm, Christian Glaser, Naser Damer, Arjan Kuijper, Alexander Nouak, Jukka Komulainen, Tiago Freitas Pereira, Shubham Gupta, Shubham Khandelwal, Shubham Bansal, Ayush Rai, Tarun Krishna, Dushyant Goyal, Muhammad-Adeel Waris, Honglei Zhang, Iftikhar Ahmad, Serkan Kiranyaz, Moncef Gabbouj, Roberto Tronci, Maurizio Pili, Nicola Sirena, Fabio Roli, Javier Galbally, Julian Fiérrez, Allan da Silva Pinto, Hélio Pedrini, W. S. Schwartz, Anderson Rocha, André Anjos, Sébastien Marcel: The 2nd competition on counter measures to 2D face spoofing attacks. ICB 2013: 1-6
In case of identical qualifications, preference will be given to severely disabled candidates.
The Fraunhofer-Gesellschaft is committed to providing equal career opportunities for men and women.
Fraunhofer is Europe’s largest application-oriented research organization. Our research efforts are geared entirely to people’s needs: health, security, communication, energy and the environment. As a result, the work undertaken by our researchers and developers has a significant impact on people’s lives. We are creative. We shape technology. We design products. We improve methods and techniques. We open up new vistas.
Ansprechpartner: Naser Damer