• BCI Workshop
• 
  
• Date: June 13th, 2023   Time: 9:30-12:45
• 
  
• @ "Sala Multimediale", Dipartimento di Ingegneria Industriale, Elettronica e Meccanica.
• 
  
• The event is technically sponsored by the IEEE Italy Section Biometric Council.
• 
  
• Abstract: Brain-computer interface technology holds immense potential for numerous fields and applications. Slobodan will present some BCI basics as a theoretical background for the hands-on part. The practical part of the session will address a motor imagery task using the recoveriX system for neurorehabilitation, an attention task using the "Unicorn Speller" communication tool and a code-based VEP using the "Unicorn Maze" game, which are the most frequently used techniques in the BCI. Slobodan will also show the design of closed-loop circuits with MATLAB Simulink, which perfectly illustrates BCI systems.
  
• 
  
• 9:30 - 10:15 Presentation of BCI basics, short presentation which provide theory basics
• 10:15 - 11:00 Motor imagery task using recoveriX system or biosignals recording in Flight Simulator (free online game); hands on (one volunteer is needed to play the subject - short hair one preferably)
• 11:00 - 11:30 Attention task (P300) using Unicorn Speller; hands-on exercise
• 11:30 - 12:00 Code based visual evoked potentials using Unicorn Maze game
• 12:00 - 12:45 HISYS Simulink modelling for BCI applications and closed-loop; presentation using MATLAB Simulink, presentation-exercise with different examples of real life experiment

• We are honored to host Mikkel B. Stegmann and Gabriel E. Hine for the interesting lecture "Challenges within commercial iris recognition".
• 
  
• Date: October 6th, 2022   Time: 14:00-15:00
• 
  
• @ "Sala Multimediale", Dipartimento di Ingegneria Industriale, Elettronica e Meccanica.
• 
  
• The event is technically sponsored by the IEEE Italy Section Biometric Council.
• 
  
• Abstract: In this talk, we will introduce the biometric company Fingerprint Cards and discuss high-level interests within commercial iris recognition. Challenges specific to the iris modality will be reviewed, and comparisons to competing modalities such as face recognition will be made. We will end by introducing the audience to the classic Daugman iris recognition pipeline and mention collaborative possibilities.
  
• 
  
• Bio: Mikkel B. Stegmann received the received the Ph.D. degree in applied mathematics from the Technical University of Denmark in 2004. Following a post.doc. on machine learning applied to medical images, he moved to 3Shape A/S to conduct design, principal architecture and management of 3D laser scanners during the years 2006-2011. His subsequent four years were spent with Teklatech A/S modelling ASIC power delivery networks and later, in the role of Principal Scientist, he pioneered a unique, constrained, multi-objective optimization engine. Since 2015 he has been creating biometric technology with Fingerprint Cards A/S, currently specializing in iris recognition and face analysis as Director of the Touchless Technologies unit. His technical interests include optimization, machine learning, biometry, system architecture, cameras / optics / illuminators, software design and edge computing.
• Bio: Gabriel E. Hine received his Ph.D. in Applied Electronics from Roma Tre University in 2019 with the thesis ''Anonymous biometrics''. Since 2020 he has been working with Fingerprint Cards A/S as an algorithms and systems engineer, where he has filed four patents on liveness detection. His technical and research interests include signal processing, information security, and biometrics.

• We are honored to host Anderson Rocha for his interesting lecture "Reasoning for Complex Data through Ensemble-based Self-Supervised Learning".
• 
  
• Date: June 28th, 2022   Time: 16:00-17:00
• 
  
• @ "Sala Multimediale", Dipartimento di Ingegneria Industriale, Elettronica e Meccanica.
• 
  
• The event is technically sponsored by the IEEE Italy Section Biometric Council.
• 
  
• Abstract: Self-supervised learning deals with problems that have little or no available labeled data. Recent work has shown impressive results when underlying classes have significant semantic differences. One important dataset in which this technique thrives is ImageNet, as intra-class distances are substantially lower than inter-class distances. However, this is not the case for several critical tasks, and general self-supervised learning methods fail to learn discriminative features when classes have closer semantics, thus requiring more robust strategies. We propose a strategy to tackle this problem and to enable learning from unlabeled data even when samples from different classes are not prominently diverse. We approach the problem by leveraging a novel ensemble-based clustering strategy where clusters derived from different configurations are combined to generate a better grouping for the data samples in a fully-unsupervised way. This strategy allows clusters with different densities and higher variability to emerge, which in turn reduces intra-class discrepancies, without requiring the burden of finding an optimal configuration per dataset. We also consider different Convolutional Neural Networks to compute distances between samples. We refine these distances by performing context analysis and grouping them to capture complementary information. We consider two applications to validate our pipeline: Person Re-Identification and Text Authorship Verification. These are challenging applications considering that classes are semantically close to each other and that training and test sets have disjoint identities. Our method is robust across different modalities and outperforms state-of-the-art results with a fully-unsupervised solution without any labeling or human intervention.
  
• 
  
• Nano-bio: Anderson Rocha is a full professor of artificial intelligence and digital forensics at the Institute of Computing, University of Campinas (Unicamp), Brazil. His main interests include Artificial Intelligence, Digital Forensics, and Reasoning for Complex Data He is a Senior IEEE member, an elected affiliate member of the Brazilian Academy of Sciences (ABC) and the Brazilian Academy of Forensic Sciences (ABC). He is a two-term elected member of the IEEE Information Forensics and Security Technical Committee (IFS-TC) and served as its Chair for the 2019-2020 term. He is a Microsoft Research and a Google Research Faculty Fellow, important academic recognitions given to researchers by Microsoft Research and Google, respectively.

The paper "On the Statistical Independence of Parametric Representations in Biometric Cryptosystems: Evaluation and Improvement", authored by Riccardo Musto, Emanuele Maiorana, Ridvan S. Kuzu, Gabriel E. Hine and Patrizio Campisi, received the BEST PAPER AWARD at the 11th International Conference on Pattern Recognition Applications and Methods (ICPRAM), held from February 3 - 5, 2022.

On 15 September 2021 the European Association for Biometrics (EAB) awarded young researchers for their outstanding works in the area of biometrics. The selection for the research award has been made on the basis of the academic and scientific quality of the submitted works as well as on the quality of their presentations. An international jury chose 3 candidates out of a broad range of submitted high quality papers to present their significant contribution in front of the jury, the EAB members and the public audience in a virtual meeting.

The European Biometrics Research Award 2021 was granted to Ridvan Salih Kuzu from Roma Tre University (Italy) for his PhD-thesis ''On-the-fly Vascular Multibiometrics''. With his PhD work releases the constraint of fixed hand in touchless vein biometrics and enables walk-through-style biometrics. The thesis proves that low-cost sensors are sufficient for robust vein biometrics if multi-exposure images are exploited and customized deep neural network architectures are applied. Moreover the PhD work demonstrated that soft biometric traits like gender or age are reliable information to enhance the security of vein biometrics.