Stefano Ricciardi

Ubiquitous iris recognition by means of mobile devices

Iris authentication/recognition on mobile devices is feasible.Spatial histograms can be exploited for iris features extraction and matching.Performance of iris segmentation/recognition algorithms is strongly affected by capture conditions.Imaging sensors resolution alone does not necessarily result in higher recognition accuracy. The worldwide diffusion of latest generations mobile devices, namely smartphones and tablets, represents the technological premise to a new wave of applications for which reliable owner identification is becoming a key requirement. This crucial task can be approached by means of biometrics (face, iris or fingerprint) by exploiting high resolution imaging sensors typically built-in on this class of devices, possibly resulting in a ubiquitous platform to verify owner identity during any kind of transaction involving the exchange of sensible data. Among the aforementioned biometrics, iris is known for its inherent invariance and accuracy, though only a few works have explored this topic on mobile devices. In this paper a comprehensive method for iris authentication on mobiles by means of spatial histograms is described. The proposed approach has been tested on the MICHE-I iris dataset, featuring subjects captured indoor and outdoor under controlled and uncontrolled conditions by means of built-in cameras aboard three among the most diffused smartphones/tablets on the market. The experimental results collected, provide an interesting insight about the readiness of mobile technology with regard to iris recognition.

A haptic-based approach to virtual training for aerospace industry

In the last years, the industrial world has been increasingly adopting computer-aided solutions for design for maintainability and maintenance training tasks with the goal to reduce development costs and to shorten time, and to improve product and service quality. Computer-based training systems created to simulate machine assembly maintenance are normally operated by means of ordinary human-computer interfaces (keyboard, mouse, etc.), but this usually results in systems that are far from the real procedures, and therefore not effective in terms of training. In this study, we show that a better solution may come from the combination of virtual reality techniques and haptic interaction. To this regard, we present the results of a research aimed at testing and evaluating the effectiveness of the haptic feedback for first-person maintenance tasks targeted to the aerospace industry. The proposed system implements an interaction environment in which each of the main maintenance activities can be simulated by the trainee exploiting a hand-based commercial haptic device, operated by means of specific haptic-rendering techniques to provide realistic feedbacks during manipulation. A usability study is included to help assessing the potential of this approach.

Ambient intelligence framework for context aware adaptive applications

Despite recent turbulence of the digital economy, the information society continues its progress. Information and communication technologies (ICT) are increasingly entering in all aspects of our life and in all sectors, opening a world of unprecedented scenarios where people interact with electronic devices embedded in environments that are sensitive and responsive to the presence of users. These context-aware environments combine ubiquitous information, communication, with enhanced personalization, natural interaction and intelligence. A critical issue, common in most of applications framed inside domotic systems or ambient intelligence, is the approach to automatically detect context from wearable or environmental sensor systems and to transform such information for achieving personalized and adaptive services. Most of the flexible and robust systems use probabilistic detection algorithms that require extensive libraries of training; in this work we experiment with a prototype framework based on intelligent agents skilled to capture user habits, identify requests, and apply the artefact-mediated activity through hybrid approaches, featuring adaptive fuzzy control strategy and biometric techniques.

Iris recognition through machine learning techniques: A survey ☆

Abstract Iris recognition is one of the most promising fields in biometrics. Notwithstanding this, there are not so many research works addressing it by machine learning techniques. In this survey, we especially focus on recognition, and leave the detection and feature extraction problems in the background. However, the kind of features used to code the iris pattern may significantly influence the complexity of the methods and their performance. In other words, complexity affects learning, and iris patterns require relatively complex feature vectors, even if their size can be optimized. A cross-comparison of these two parameters, feature complexity vs. learning effectiveness, in the context of different learning algorithms, would require an unbiased common benchmark. Moreover, at present it is still very difficult to reproduce techniques and experiments due to the lack of either sufficient implementation details or reliable shared code.

Fast 3D face recognition based on normal map

This paper presents a 3D face recognition method aimed to biometric applications. The proposed method compares any two faces represented as 3D polygonal surfaces through their corresponding normal map, a bidimensional array which stores local curvature (mesh normals) as the pixels RGB components of a color image. The recognition approach, based on the computation of a difference map resulting from the comparison of normal maps, is simple yet fast and accurate. A weighting mask, automatically generated for each subject using a set of expression variations, improves the robustness to a broad range of facial expressions. First results show the effectiveness of the method on a database of 3D faces featuring different genders, ages and expressions.

FACES: 3D FAcial reConstruction from anciEnt Skulls using content based image retrieval

Abstract Powerful techniques for modelling and rendering tridimensional organic shapes, like human body, are today available for applications in many fields such as special effects, ergonomic simulation or medical visualization, just to name a few. These techniques, combined with Content Based Image Retrieval (CBIR), are proving to be very useful also to archaeologists and anthropologists committed to reconstruct the aspect of the inhabitants of historically relevant sites like Pompei. This paper presents an integrated system to provide 3D FAcial reConstruction from anciEnt Skulls (FACES). FACES, starting from radiological analysis of an ancient skull and a database of modern individuals of the same area/gender/age, produces a tridimensional facial model compatible to the anthropological and craniometrical features of the original skull. Finally, we compare FACES peculiarities to the most used facial reconstruction methodologies available today.

White Paper on Industrial Applications of Computer Vision and Pattern Recognition

The paper provides a summary of the contributions to the industrial session at ICIAP2013, describing a few practical applications of Video Analy- sis, in the Surveillance and Security field. The session has been organized to stimulate an open discussion within the scientific community of CVPR on new emerging research areas which deserve particular attention, and may contribute to the improvement of industrial applications in the near future.

Fast Iris Recognition on Smartphone by means of Spatial Histograms

The iris has been proposed as a highly reliable and stable biometric identifier for person authentication/recognition about two decades ago. Since then, most work in the field has been focused on segmentation and matching algorithms able to work on pictures of whole face or eye region typically captured at close distance, while preserving recognition accuracy. In this paper we present an iris matching algorithm based on spatial histograms that, while showing good recognition performance on some of the most referenced public iris dataset, is also able to perform a one-to-one comparison in a small amount of time thanks to its low computing load, thus resulting particularly suited to iris recognition applications on mobile devices.

Dependability issues in visual-haptic interfaces

Dependability of a system is commonly referred to its reliability, its availability and its maintenability (RAM), but when this concept is applied to user interfaces there is no common agreement on what aspects of user-system interaction are related to a satisfactory RAM level for the whole system. In particular, when dealing with haptic systems, interface dependability may become a crucial issue in medical and in military domains when life-critical systems are to be manipulated or where costly remote control operations are to be performed, like in industrial processes control or in aerospace/automotive engineering and manufacturing. This paper discusses the role of dependability in haptic user interfaces, aiming to the definition of a framework for the assessment of the usability and dependability properties of haptic systems and their possible correlations. The research is based on the analysis of a visual-haptic-based simulator targeted to maintenance activity training for aerospace industry which is taken as a case study. As a result, we propose a novel framework able to collect and then process relevant interaction data during the execution of haptic tasks, enabling to analyze dependability vs. usability correlations.

A Pervasive Visual–Haptic Framework for Virtual Delivery Training

Thanks to the advances of voltage regulator (VR) technologies and haptic systems, virtual simulators are increasingly becoming a viable alternative to physical simulators in medicine and surgery, though many challenges still remain. In this study, a pervasive visual-haptic framework aimed to the training of obstetricians and midwives to vaginal delivery is described. The haptic feedback is provided by means of two hand-based haptic devices able to reproduce force-feedbacks on fingers and arms, thus enabling a much more realistic manipulation respect to stilus-based solutions. The interactive simulation is not solely driven by an approximated model of complex forces and physical constraints but, instead, is approached by a formal modeling of the whole labor and of the assistance/intervention procedures performed by means of a timed automata network and applied to a parametrical 3-D model of the anatomy, able to mimic a wide range of configurations. This novel methodology is able to represent not only the sequence of the main events associated to either a spontaneous or to an operative childbirth process, but also to help in validating the manual intervention as the actions performed by the user during the simulation are evaluated according to established medical guidelines. A discussion on the first results as well as on the challenges still unaddressed is included.