410 research outputs found
Fingerprint verification competition 2006
No full textThe interest in fingerprint-based biometric systems has constantly grown in recent years and considerable efforts have been focused by both academia and industry on the development of new algorithms for fingerprint recognition. Raffaele Cappelli, Matteo Ferrara, Annalisa Franco and Davide Maltoni of the Biometric System Laboratory at the University of Bologna explain the findings of the Fingerprint Verification Competition 2006
Fingerprint Synthesis
No full textBook cover
Handbook of Fingerprint Recognition pp 385–426Cite as
Fingerprint Synthesis
Davide Maltoni, Dario Maio, Anil K. Jain & Jianjiang Feng
Chapter
First Online: 05 July 2022
253 Accesses
Abstract
Synthetic fingerprints, when properly generated, represent a reasonable substitute for real fingerprints for the design, training, and benchmarking of fingerprint recognition algorithms. This approach is particularly useful to deal with emerging privacy regulations (e.g., EU-GDPR) limiting the use of personally identifiable information. This chapter introduces fingerprint synthesis and focuses on the two main categories of generation approaches: (i) first generate a master fingerprint and then derive multiple impressions (e.g., SFinGe); (ii) generative models (e.g., GAN) for the direct synthesis of fingerprint images. Validation of synthetic generators through large scale experiments is finally presented
Enhancing decision-making with data-driven insights in critical situations: impact and implications of AI-powered predictive solutions
No full textIdentifying early signals of crisis or insolvency is essential, for firms, to enable timely
interventions, preserve financial health, and sustain competitive advantage. Traditional financial
models, while foundational in this domain, often fall short in complex environments and there is a
widening gap between the theoretical advancements in predictive modeling and their practical
application in real-world decision-making, particularly for small and medium-sized enterprises, that
may lack the resources to implement sophisticated risk management tools. Considering the Italian
business landscape, this research examines the integration of advanced AI-driven techniques,
showcasing their ability to handle vast and diverse data sources while improving predictive accuracy.
By employing an inductive, multi-layered approach to model implementation and refinement, the
study demonstrates how AI-based models can identify early indicators of distress up to five years before
insolvency, thus providing a robust framework for preventive strategies. A key added value is the focus
on enhancing explainability and understandability within AI models, delivering critical insights for
both academia and managers; by improving transparency, these models not only strengthen predictive
capacity but also enable stakeholders to better understand and interpret the drivers of financial and
economic distress.
Furthermore, the study highlights the challenges of adopting new technologies in
organizational contexts, addressing issues such as data ethics, managerial literacy, over-reliance risks
and the alignment of AI-based decision tools with regulatory standards. The findings contribute both
to academic discourse and practical applications by advocating for a cultural shift toward data-driven
decision-making in critical phases of business life; by balancing the precision of Machine Learning with
the interpretability required for managerial planning, it is underscored that, while AI just complements
human expertise, it is, today, instrumental in surviving in complex economic landscapes with greater
foresight and resilience
On the Feasibility of Creating Double-Identity Fingerprints
No full textA double-identity fingerprint is a fake fingerprint created by combining features from two different fingers, so that it has a high chance to be falsely matched with fingerprints from both fingers. This paper studies the feasibility of creating double-identity fingerprints by proposing two possible techniques and evaluating to what extent they may be used to fool the state-of-the-art fingerprint recognition systems. The results of systematic experiments suggest that existing algorithms are highly vulnerable to this specific attack (about 90% chance of success at FAR = 0.1%) and that the fingerprint patterns generated might be realistic enough to fool human examiners
Large-scale fingerprint identification on GPU
No full textThis paper proposes a new parallel algorithm to speed up fingerprint identification using GPUs. A careful design of the algorithm and data structures, guided by well-defined optimization goals, yields a speed-up of 1946× over a baseline sequential CPU implementation and of 207× over a CPU implementation optimized with SIMD instructions. The proposed algorithm enables a medium-scale AFIS (Automated Fingerprint Identification System) to run on a simple PC with four Tesla C2075 GPUs. On a benchmark with 250 000 fingerprints and 100 000 queries, the proposed system yields state-of-the-art biometric accuracy with a throughput of more than 35 million fingerprint matches per second. The proposed approach can be easily scaled-up, thus making possible the implementation of a large-scale AFIS (i.e., with a database of hundred million fingerprints) on inexpensive hardware. © 2015 Elsevier Inc. All rights reserved
Large scale fingerprint recognition accelerated in hardware
No full textTo make automatic fingerprint identification systems (AFIS) capable of searching across several millions of fingerprints in a few seconds, very powerful (and expensive) distributed computing architectures are typically used. The recent improvement of algorithms and the availability of powerful CPUs and GPUs makes it now possible to deploy large scale fingerprint recognition on low-cost hardware, thus approaching a larger number of applications (e.g., welfare benefits in poor countries). This chapter discusses architectural design, algorithms and hardware optimization to speed-up fingerprint recognition on large databases
Generating synthetic fingerprints
No full textSynthetic Fingerprint generation techniques and associated tools (e.g., SFinGe) were introduced more than 15 years ago. The main aim was to generate large databases for performance evaluation without allocating huge amount of resources for acquisition campaigns and, at the same time, to conform with the privacy directives that in many countries limit the exchange of biometric data. While the original scope remains central today, since the generation of very large synthetic dataset is crucial to predict accuracy on very-large scenarios, new security needs (such as detecting altered fingerprints) and algorithms improvements (supervised learning approaches) are continuously renewing interest in the generation of synthetic fingerprints
Pontano e la guerra: il “De bello Neapolitano” nel suo contesto storico, ideologico e letterario
No full textMonographic section of Discussions, which, in this first issue, collects papers by Francesco Storti, Davide Morra, Fulvio Delle Donne, Guido Cappelli and Antonietta Iacono.Sezione monografica di Confronti, che, in questo primo numero, raccoglie i contributi di Francesco Storti, Davide Morra, Fulvio Delle Donne, Guido Cappelli e Antonietta Iacono
IN-service inspection and on-line monitoring of inaccessible components in nuclear power plants using guided wave technology
No full textOne of the most challenging problems in the on-line monitoring of critical parameters of nuclear plants is the inspection of components that result inaccessible or difficult to reach. In this context, there is an increasing interest of the scientific community and industry for the use of Ultrasonic Guided Waves (UGW) for addressing this issue. In this work, the problem of the applicability of the UGW technique with magnetostrictive sensors to NPP structures is described, together with the outline of the related advantages as well as the main technical concerns that may arise from such applications. This methodology has been tested on experimental activities concerning high temperature applications. Results show the effectiveness of such an approach
Determining the authenticity of PDO buffalo mozzarella: an approach based on Fourier transform infrared (MIR-FTIR) spectroscopy and on chemometric tools
No full textThe potential of Mid-infrared spectroscopy coupled with chemometric tools was evaluated for the authentication and discrimination of PDO (Protected Denomination of Origin) buffalo mozzarella produced by traditional and industrial cheese-making processes.
Samples of mozzarella provided by local producers and supermarkets were analysed through both official destructive methods and Attenuated Total Reflectance-Fourier transform infrared spectroscopy (FTIR/ATR). In particular, destructive methods allowed to determine the content of fatty substances, proteins, moisture and total nitrogen.
The results show that only the conjunction of MID-infrared spectroscopy with chemometric analysis can provide a satisfying solution to discriminate between the different types of mozzarella
- …
