Abstract In conventional machine learning (ML), a fundamental assumption is that the training and test sets share identical feature distributions, a reasonable premise drawn from the same dataset. However, real-world scenarios often defy this assumption, as data may originate from diverse sources, causing disparities between training and test data distributions. This leads to a domain shift, where variations emerge between the source and target domains. This study delves into human action recognition (HAR) models within an unconstrained, real-world setting, scrutinizing the impact of input data variations related to contextual information and video encoding. The objective is to highlight the intricacies of model performance and interpretability in this context. Additionally, the study explores the domain adaptability of HAR models, specifically focusing on their potential for re-identifying individuals within uncontrolled environments. The experiments involve seven pre-trained backbone models and introduce a novel analytical approach by linking domain-related (HAR) and domain-unrelated (re-identification (re-ID)) tasks. Two key analyses addressing contextual information and encoding strategies reveal that maintaining the same encoding approach during training results in high task correlation while incorporating richer contextual information enhances performance. A notable outcome of this study is the comprehensive evaluation of a novel transformer-based architecture driven by a HAR backbone, which achieves a robust re-ID performance superior to state-of-the-art (SOTA). However, it faces challenges when other encoding schemes are applied, highlighting the role of the HAR classifier in performance variations.

VQAsk is an Android application that helps visually impaired users to get information about images framed by their smartphones. It enables to interact with one’s photographs or the surrounding visual environment through a question-and-answer interface integrating three modalities: speech interaction, haptic feedback that facilitates navigation and interaction, and sight. VQAsk is primarily designed to help visually impaired users mentally visualize what they cannot see, but it can also accommodate users with varying levels of visual ability. To this aim, it embeds advanced NLP and Computer Vision techniques to answer all user questions about the image on the cell screen. Image processing is enhanced by background removal through advanced segmentation models that identify important image elements. The outcomes of a testing phase confirmed the importance of this project as a first attempt at using AI-supported multimodality to enhance visually impaired users’ experience.

The use of face biometrics to automatically recognize people, though fascinating, often raises ethical concerns related to the composition of the datasets used for performance evaluation of the recognition approaches and the bias that stems from the possible demographic imbalance of age, gender, and ethnicity classes during the training phase. This study tackles such imbalance in face datasets and proposes an approach to fair age, gender, and ethnicity classification by training on finely rebalanced cohorts of face images. Special attention is devoted to the ethical aspects related to having face samples of real people vs. having synthetic face samples (generated with a Generative Adversarial Network - GAN - model). Therefore, the dataset rebalancing approach exploits synthetic images instead of real ones in order to decrease the possible privacy concerns raised by new image captures. The work further aims to demonstrate that gross re-balancing is insufficient to solve all the problems related to fair demographic classification, but a finer strategy is worth adopting. The experiments compare rebalancing the single demographic classes with a finer strategy considering classes characterized by combinations of features. This entails analyzing the imbalance of the different cohorts in the dataset and appropriately rebalancing them to evaluate the new performance.

The popularity of biometrics-based user identification has significantly increased over the last few years. User identification based on the face, fingerprints, and iris, usually achieves very high accuracy only in controlled setups and can be vulnerable to presentation attacks, spoofing, and forgeries. To overcome these issues, this work proposes a novel strategy based on a relatively less explored biometric trait, i.e., gait, collected by a smartphone accelerometer, which can be more robust to the attacks mentioned above. According to the wearable sensor-based gait recognition state-of-the-art, two main classes of approaches exist: 1) those based on machine and deep learning; 2) those exploiting hand-crafted features. While the former approaches can reach a higher accuracy, they suffer from problems like, e.g., performing poorly outside the training data, i.e., lack of generalizability. This paper proposes an algorithm based on hand-crafted features for gait recognition that can outperform the existing machine and deep learning approaches. It leverages a modified Majority Voting scheme applied to Fast Window Dynamic Time Warping, a modified version of the Dynamic Time Warping (DTW) algorithm with relaxed constraints and majority voting, to recognize gait patterns. We tested our approach named MV-FWDTW on the ZJU-gaitacc, one of the most extensive datasets for the number of subjects, but especially for the number of walks per subject and walk lengths. Results set a new state-of-the-art gait recognition rate of 98.82% in a cross-session experimental setup. We also confirm the quality of the proposed method using a subset of the OU-ISIR dataset, another large state-of-the-art benchmark with more subjects but much shorter walk signals.

This work deals with the delicate task of lie detection from facial dynamics. The proposed Face Truth Machine (FTM) is an intelligent system able to support a human operator without any special equipment. It can be embedded in the present infrastructures for forensic investigation or whenever it is required to assess the trustworthiness of responses during an interview. Due to its flexibility and its non-invasiveness, it can overcome some limitations of present solutions. Of course, privacy issues may arise from the use of such systems, as often underlined nowadays. However, it is up to the utilizer to take these into account and make fair use of tools of this kind. The paper will discuss particular aspects of the dynamic analysis of face landmarks to detect lies. In particular, it will delve into the behavior of the features used for detection and how these influence the system’s final decision. The novel detection system underlying the Face Truth Machine is able to analyze the subject’s expressions in a wide range of poses. The results of the experiments presented testify to the potential of the proposed approach and also highlight the very good results obtained in cross-dataset testing, which usually represents a challenge for other approaches. • The paper presents an approach to lie detection based on face micro-expressions. • The method uses hand-crafted features to spot events suggested by psychological investigations. • Thresholds on facial micro-expressions support detecting the occurrence of relevant events. • The transfer of parameters across datasets achieved satisfying results. • The contribution of single facial features is analyzed via decidability indices.

The aim of this paper is to investigate emotion recognition using a multimodal approach that exploits convolutional neural networks (CNNs) with multiple input. Multimodal approaches allow different modalities to cooperate in order to achieve generally better performances because different features are extracted from different pieces of information. In this work, the facial frames, the optical flow computed from consecutive facial frames, and the Mel Spectrograms (from the word melody) are extracted from videos and combined together in different ways to understand which modality combination works better. Several experiments are run on the models by first considering one modality at a time so that good accuracy results are found on each modality. Afterward, the models are concatenated to create a final model that allows multiple inputs. For the experiments the datasets used are BAUM-1 ((Bahçeşehir University Multimodal Affective Database - 1) and RAVDESS (Ryerson Audio–Visual Database of Emotional Speech and Song), which both collect two distinguished sets of videos based on the different intensity of the expression, that is acted/strong or spontaneous/normal, providing the representations of the following emotional states that will be taken into consideration: angry, disgust, fearful, happy and sad. The performances of the proposed models are shown through accuracy results and some confusion matrices, demonstrating better accuracy than the compared proposals in the literature. The best accuracy achieved on BAUM-1 dataset is about 95%, while on RAVDESS it is about 95.5%.

The outbreak of digital devices on the Internet, the exponential diffusion of data (images, video, audio, and text), along with their manipulation/generation also by artificial intelligence models, such as generative adversarial networks, have created a great deal of concern in the field of forensics. A malicious use can affect relevant application domains, which often include counterfeiting biomedical images and deceiving biometric authentication systems, as well as their use in scientific publications, in the political world, and even in school activities. It has been demonstrated that manipulated pictures most likely represent indications of malicious behavior, such as photos of minors to promote child prostitution or false political statements. Following this widespread behavior, various forensic techniques have been proposed in the scientific literature over time both to defeat these spoofing attacks as well as to guarantee the integrity of the information. Focusing on image forensics, which is currently a very hot topic area in multimedia forensics, this article will present the whole scenario in which a target image could be modified. The aim of this comprehensive survey will be (1) to provide an overview of the types of attacks and contrasting techniques and (2) to evaluate to what extent the former can deceive prevention methods and the latter can identify counterfeit images. The results of this study highlight how forgery detection techniques, sometimes limited to a single type of real scenario, are not able to provide exhaustive countermeasures and could/should therefore be combined. Currently, the use of neural networks, such as convolutional neural networks, is already heading, synergistically, in this direction.

Emotion estimation from face expression analysis is nowadays a widely-explored computer vision task. In turn, the classification of expressions relies on relevant facial features and their dynamics. Despite the promising accuracy results achieved in controlled and favorable conditions, the processing of faces acquired at a distance, entailing low-quality images, still suffers from a significant performance decrease. In particular, most approaches and related computational models become extremely unstable in the case of the very small amount of useful pixels that is typical in these conditions. Therefore, their behavior should be investigated more carefully. On the other hand, real-time emotion recognition at a distance may play a critical role in smart video surveillance, especially when controlling particular kinds of events, e.g., political meetings, to try to prevent adverse actions. This work compares facial expression recognition at a distance by: 1) a deep learning architecture based on state-of-the-art (SOTA) proposals, which exploits the whole images to autonomously learn the relevant embeddings; 2) a machine learning approach that relies on hand-crafted features, namely the facial landmarks preliminarily extracted using the popular Mediapipe framework. Instead of using either the complete sequence of frames or only the final still image of the expression, like current SOTA approaches, the two proposed methods are designed to use rich temporal information to identify three different stages of emotion. Expressions are time-split accordingly into four phases to better exploit their temporal-dependent dynamics. Experiments were conducted on the popular Extended Cohn-Kanade dataset (CK+). It was chosen for its wide use in related literature, and because it includes videos of facial expressions and not only still images. The results show that the approach relying on machine learning via hand-crafted features is more suitable for classifying the initial phases of the expression and does not decay in terms of accuracy when images are at a distance (only 0.08% of decay). On the contrary, deep learning not only has difficulties classifying the initial phases of the expressions but also suffers from relevant performance decay when considering images at a distance (52.68% accuracy decay).

Smartphones contain personal and private data to be protected, such as everyday communications or bank accounts. Several biometric techniques have been developed to unlock smartphones, among which ear biometrics represents a natural and promising opportunity even though the ear can be used in other biometric and multi-biometric applications. A problem in generalizing research results to real-world applications is that the available ear datasets present different characteristics and some bias. This paper stems from a study about the effect of mixing multiple datasets during the training of an ear recognition system. The main contribution is the evaluation of a robust pipeline that learns to combine data from different sources and highlights the importance of pre-training encoders on auxiliary tasks. The reported experiments exploit eight diverse training datasets to demonstrate the generalization capabilities of the proposed approach. Performance evaluation includes testing with collections not seen during training and assessing zero-shot cross-dataset transfer. The results confirm that mixing different sources provides an insightful perspective on the datasets and competitive results with some existing benchmarks.

This work proposes the preliminary results of a system to recognize deception from facial micro-expressions. The adopted method starts from facial landmarks to analyze the micro-dynamics underlying the facial modifications. The procedure is repeated both while answering neutral questions (entailing no reason to lie) and when answering questions whose responses may cause harm or embarrassment, at least. The preliminary results invite us to continue the research of suitable features to tackle the problem of deception detection.

One of the present approaches to gait recognition exploits the signals captured by wearable sensors, especially the accelerometers embedded in modern smartphones. However, the different speed, the ground slope, or simply the time lapse between captures cause variations that negatively affect long term recognition in a dramatic way. The proposed procedure aims at extracting gait characteristics that are as invariant as possible, and therefore useful for accurate long term recognition. The experiments compare the performance of the proposal with others in state-of-the-art that use the same benchmark, namely the ZJU-gaitacc dataset. This dataset includes a high number of samples per subject, captured in two time-separated sessions. This allows to assess the performance of the proposed method also in the long term, i.e., when comparing templates captured in different times. Most works using the same benchmark so far have not exploited both sessions. They use samples captured in the same time, constraining the use of this trait to continuous recognition, e.g., of the smartphone owner. The obtained results testify that, in this condition, the proposed feature-based method outperforms competitors in the current literature. The experiments also compare the results from a session-based partition with those obtained from a training that mixes-up samples from different sessions. As expected, the latter strategy can dramatically improve the measured performance. The significantly different results seem to suggest that the session-based partition, when feasible, can provide more realistic results, closer to the real-world application context when behavioural traits are involved in the medium/long term. The same results seem also to testify that there is still need to improve the accuracy of gait recognition via wearable sensors. This calls for further investigation of the problems related to the variability over time in the pattern of individual gait signals.

Abstract According to the Wall Street Journal, one billion surveillance cameras will be deployed around the world by 2021. This amount of information can be hardly managed by humans. Using a Inflated 3D ConvNet as backbone, this paper introduces a novel automatic violence detection approach that outperforms state-of-the-art existing proposals. Most of those proposals consider a pre-processing step to only focus on some regions of interest in the scene, i.e., those actually containing a human subject. In this regard, this paper also reports the results of an extensive analysis on whether and how the context can affect or not the adopted classifier performance. The experiments show that context-free footage yields substantial deterioration of the classifier performance (2% to 5%) on publicly available datasets. However, they also demonstrate that performance stabilizes in context-free settings, no matter the level of context restriction applied. Finally, a cross-dataset experiment investigates the generalizability of results obtained in a single-collection experiment (same dataset used for training and testing) to cross-collection settings (different datasets used for training and testing).

This work proposes BowlingVR, an advanced Virtual Reality (VR) multiplayer game that tackles two main goals: the first one is to provide a realistic User eXperience (UX) to the user, by reproducing the dynamics and physical context of a real bowling challenge; the second one is to allow a remote, distributed, socially satisfying gameplay, providing the user the illusion of the real presence of the remote players. The prototype was evaluated using a modified version of SUXES, a kind of user interview schema that was originally devised for multimedia applications and that has been modified in order to better compare the responses of different users and get a more reliable estimation of user appreciation.

Head pose estimation is a sensitive topic in video surveillance/smart ambient scenarios since head rotations can hide/distort discriminative features of the face. Face recognition would often tackle the problem of video frames where subjects appear in poses making it quite impossible. In this respect, the selection of the frames with the best face orientation can allow triggering recognition only on these, therefore decreasing the possibility of errors. This paper proposes a novel approach to head pose estimation for smart cities and video surveillance scenarios, aiming at this goal. The method relies on a cascade of two models: the first one predicts the positions of 68 well-known face landmarks; the second one applies a web-shaped model over the detected landmarks, to associate each of them to a specific face sector. The method can work on detected faces at a reasonable distance and with a resolution that is supported by several present devices. Results of experiments executed over some classical pose estimation benchmarks, namely Point '04, Biwi, and AFLW datasets show good performance in terms of both pose estimation and computing time. Further results refer to noisy images that are typical of the addressed settings. Finally, examples demonstrate the selection of the best frames from videos captured in video surveillance conditions.

The results of spoofing detection systems proposed during ASVspoof Challenges 2015 and 2017 confirmed the perspective in detection of unforseen spoofing trials in microphone channel. However, telephone channel presents much more challenging conditions for spoofing detection, due to limited bandwidth, various coding standards and channel effects. Research on the topic has thus far only made use of program codecs and other telephone channel emulations. Such emulations does not quite match the real telephone spoofing attacks. In order to asses spoofing detection methods in real scenario we present the PHONESPOOF dataset - spoofing data collected through realistic telephone channels. The PHONE-SPOOF data collection represents most threatening types of spoofing attacks and is publicly available dataset1. This work2aimed to investigate robustness of the state-of-the-art deep learning based antispoofing systems under telephone spoofing attacks conditions based on the PHONESPOOF data. Moreover newly collected dataset makes it possible to analize language dependency issue for the Anti-Spoofing methods. In the work we also focused on the development of a unified LCNN-based approach for spoofing attack detection. The goal was to train a single system able to detect various types of spoofing attacks in telephone channel. The obtained results approve the effectiveness of such solution.

The analysis of the shape and geometry of the human hand has long represented an attractive field of research to address the needs of digital image forensics. Over recent years, it has also turned out to be effective in biometrics, where several innovative research lines are pursued. Given the widespread diffusion of mobile and portable devices, the possibility of checking the owner identity and controlling the access to the device by the hand image looks particularly attractive and less intrusive than other biometric traits. This encourages new research to tacklethe present limitations. The proposed work implements the complete architecture of a mobile hand recognition system, which uses the camera of the mobile device for the acquisition of the hand in visible light spectrum. The segmentation of the hand starts from the detection of the convexities and concavities defined by the fingers, and allows extracting 57 different features from the hand shape. The main contributions of the paper develop along two directions. First, dimensionality reduction methods are investigated, in order to identify subsets of features including only the most discriminating and robust ones. Second, different matching strategies are compared. The proposed method is tested over a dataset of hands from 100 subjects. The best obtained Equal Error Rate is 0.52%. Results demonstrate that discarding features that are more prone to distortions allows lighter processing, but also produces better performance than using the full set of features. This confirms the feasibility of such an approach on mobile devices, and further suggests to adopt it even in more traditional settings.

This short paper presents the second international workshop on Games-Human Interaction - GHItaly 2018. The goal of this series of workshops is to focus on advanced aspects of the design and development of game interfaces. The quality of the resulting interaction is a highly relevant issue for creating an engaging and satisfactory user experience, especially when deeply multidimensional artefacts such as video games are concerned.

Context awareness is major component of Ambient Intelligence. In fact, Ambient Intelligent environments are designed to combine ubiquity, awareness, intelligence and natural interaction. Awareness is defined as the ability by the system to locate and recognize people and objects, and their intentions. Then, intelligence is the ability of the system to analyze the detected context and to adapt its behavior to people and situations, and to learn over time, in order to provide users with personalized services. These concepts date back to late'90, but nowadays the widespread and ubiquitous availability of mobile devices, equipped with several different sensors, allows to put them into practice in a number of unexpected ways. This work presents a preliminary investigation on the possibility to use some of the smartphone sensors, namely the accelerometer and the gyroscope, to identify the bodily context when the user lifts the device to answer a call. The arm gesture, i.e., the way it is performed, is classified into 4 different states: while standing, sitting, walking or running. This information can be used to trigger context-sensitive system actions.

Gait recognition has been traditionally tackled by computer vision techniques. As a matter of fact, this is a still very active research field. More recently, the spreading use of smart mobile devices with embedded sensors has also spurred the interest of the research community for alternative methods based on the gait dynamics captured by those sensors. In particular, signals from the accelerometer seem to be the most suited for recognizing the identity of the subject carrying the mobile device. Different approaches have been investigated to achieve a sufficient recognition ability. This paper proposes an automatic extraction of the most relevant features computed from the three raw accelerometer signals (one for each axis). It also presents the results of comparing this approach with a plain Dynamic Time Warping (DTW) matching. The latter is computationally more demanding, and this is to take into account when considering the resources of a mobile device. Moreover, though being a kind of basic approach, it is still used in literature due to the possibility to easily implement it even directly on mobile platforms, which are the new frontier of biometric recognition.

Here we discuss a framework (OpenAnswer) providing support to the teacher's activity of grading answers to open ended questions. OpenAnswer implements a teacher mediated peer-evaluation approach: the marking results obtained from peer assessments are tuned by the grades explicitly assigned by the teacher, the teacher grades only a subset of the answers, suggested by the system. When a termination criterion is met, for the process managing the amount of teacher grading work, the remaining answers are automatically graded. A Bayesian Network is designed to represent the information related to students' models, peer assessments, and teacher's grading. The model parameters are many, here we report the results of investigations on a particularly tricky aspect of the framework, that is the modeling and optimization of the Conditional Probability Tables that are an important part of the Bayesian underlying model. In fact, they express the hypothesized relation between items of information that are relevant for evidence propagation through the network. Results suggest that this optimization improves OpenAnswer's performance, i.e. its capability to infer correct grades. We also show evidence of the influence of the teacher's assessing style on the grading process.

We discuss the design and development of HCI models for authentication based on gait and gesture that can be supported by mobile and wearable equipment. The paper proposes to use such biometric behavioral traits for partially transparent and continuous authentication by means of behavioral patterns.

In this paper we experimentally investigate the influence of several factors on the final performance of an automatic grade prediction system based on teacher-mediated peer assessment. Experiments are carried out by OpenAnswer, a system designed for peer assessment of open-ended questions. It exploits a Bayesian Network to model the students' learning state and the propagation of information injected in the system by peer grades and by a (partial) grading from the teacher. The relevant variables are characterized by a probability distribution (PD) of their discrete values. We aim at analysing the influence of the initial set up of the PD of these variables on the ability of the system to predict a reliable grade for answers not yet graded by the teacher. We investigate here the influence of the initial choice of the PD for the student's knowledge (K), especially when we have no information on the class proficiency on the examined skills, and of the PD of the correctness of student's answers, conditioned by her knowledge, P(C|K). The latter is expressed through different Conditional Probability Tables (CPTs), in turn, to identify the one allowing to achieve the best final results. Moreover we test different strategies to map the final PD for the correctness (C) of an answer, namely the grade that will be returned to the student, onto a single discrete value. Copyright © 2016 by SCITEPRESS-Science and Technology Publications, Lda. All rights reserved.

The growing interest for mobile biometrics stems from the increasing need to secure personal data and services, which are often stored or accessed from there. Modern user mobile devices, with acquisition and computation resources to support related operations, are nowadays widely available. This makes this research topic very attracting and promising. Iris recognition plays a major role in this scenario. However, mobile biometrics still suffer from some hindering factors. The resolution of captured images and the computational power are not comparable to desktop systems yet. Furthermore, the acquisition setting is generally uncontrolled, with users who are not that expert to autonomously generate biometric samples of sufficient quality. Mobile Iris CHallenge Evaluation aims at providing a testbed to assess the progress of mobile iris recognition, and to evaluate the extent of its present limitations. This paper presents the results of the competition launched at the 2016 edition of the International Conference on Pattern Recognition (ICPR).

The analysis of answers to open-ended questions provides greatly accurate assessment, being in turn demanding for the teacher. Here we show an approach exploiting peer assessment to partially relieve the teacher, and to provide information on the meta-cognitive ability of students of making correct evaluations on their peers. Open Answer handles a Bayesian model for each student, representing her/his learning state and judgment capability. The students' sub-networks are connected through peer-assessment. The process end up with a full set of grades for all students' answers, after the teacher had actually graded only part of them. We present experimental data and simulations aiming at identifying the best strategies to exploit the available information.

The ability of a biometric system to reliably recognize registered individuals significantly depends on the kind and amount of variation that the exploited biometric trait may undergo throughout acquisitions. Those variations may be due both to acquisition devices, or to different environment settings, or to modification of the trait appearance. One of the strategies to address changes in biometric features is to store more templates for the same person, in order to increase the chances to identify her. The problem arises to choose the templates to store in a way which actually achieves better performance, while avoiding flooding the system with an excessively huge gallery. This work proposes an approach for the selection of the best templates for face recognition, which is based on a notion of gallery entropy, and can be also used for large datasets. Though relying on a clustering process, its main achievement is to automatically derive the best number of clusters/prototypes per subject without requiring to fixing it in advance. Comparative tests with existing approaches show that it is a very promising solution.

Advanced surveillance applications often require the ability to re-identify an individual. In the typical context of a sensor network, this means to recognize a subject acquired at one location among a feasible set of candidates acquired at another locations and/or over time. Usually this does not necessarily imply to know the identity, and actually it might remain unknown. This task is especially challenging in applications targeted at crowded environments. Face and gait are contactless biometrics which are particularly suited to be used for re-identification, but even “soft” biometrics have also been considered to this aim. This paper presents a comparison of approaches to re-identification, with some characteristic examples in literature, while a complete overview of the main researches is out of our scope. The goal is to provide an objective estimate of both the state-of-the-art and the potential of such techniques to further improve them and to extend the applicability of re-identification systems.

Multi-biometric systems are a powerful solution to deal with limitations of single classifiers, therefore improving the final recognition accuracy. The sub-systems composing the final architecture often return supplementary indices of input quality and/or of response reliability, which further qualify each recognition score. These indices can enter different information fusion policies. First, they can be used as weights for the fusion of the corresponding scores, in such a way that less trustworthy responses have a lower influence. Alternatively, they can be used to drive the selection of a subset of systems actually enabled for each fusion operation. The present work discusses their appropriate combination with respective scores, to obtain single values which are easier to handle and compare. It is worth underlining the different nature of quality and reliability measures. The quality estimation of input samples requires a complex analysis of environmental conditions, including capture sensors, besides computations over acquired data. Reliability of a system estimates its ability to return a correct response. As an alternative to combination, some solutions rather estimate the joint distributions of conditional probabilities of the scores from the single subsystems. These solutions require training through a huge number of samples. Furthermore, they assume stable score distributions. Our unified representation of the recognition score and of the corresponding quality/reliability value into a single complex number provides simplification and speed up of fusion of multi-classifier results. It also allows to devise procedures to readily compare the performance of different modules in a multi-biometric system, given that there is no natural ordering of these pairs of values of different nature. Moreover our method achieves performance comparable to top performing schemes, yet does not require a prior estimation of (joint) score distributions. As a matter of fact, though representing an upper bound to the obtainable performance, Likelihood ratio has the limit to require an accurate estimation of score distributions, while our approach relies on the reliability of each single response. This feature is very interesting when the set of relevant subjects may present significant variations over time.

Face recognition has made significant advances in the last decade, but robust commercial applications are still lacking. Current authentication/identification applications are limited to controlled settings, e.g., limited pose and illumination changes, with the user usually aware of being screened and collaborating in the process. Among others, pose and illumination changes are limited. To address challenges from looser restrictions, this paper proposes a novel framework for real-world face recognition in uncontrolled settings named Face Analysis for Commercial Entities (FACE). Its robustness comes from normalization (“correction”) strategies to address pose and illumination variations. In addition, two separate image quality indices quantitatively assess pose and illumination changes for each biometric query, before submitting it to the classifier. Samples with poor quality are possibly discarded or undergo a manual classification or, when possible, trigger a new capture. After such filter, template similarity for matching purposes is measured using a localized version of the image correlation index. Finally, FACE adopts reliability indices, which estimate the “acceptability” of the final identification decision made by the classifier. Experimental results show that the accuracy of FACE (in terms of recognition rate) compares favorably, and in some cases by significant margins, against popular face recognition methods. In particular, FACE is compared against SVM, incremental SVM, principal component analysis, incremental LDA, ICA, and hierarchical multiscale local binary pattern. Testing exploits data from different data sets: CelebrityDB, Labeled Faces in the Wild, SCface, and FERET. The face images used present variations in pose, expression, illumination, image quality, and resolution. Our experiments show the benefits of using image quality and reliability indices to enhance overall accuracy, on one side, and to provide for individualized processing of biometric probes for better decision-making purposes, on the other side. Both kinds of indices, owing to the way they are defined, can be easily integrated within different frameworks and off-the-shelf biometric applications for the following: 1) data fusion; 2) online identity management; and 3) interoperability. The results obtained by FACE witness a significant increase in accuracy when compared with the results produced by the other algorithms considered.

We present SWift (SignWriting improved fast transcriber), an advanced editor for computer-aided writing and transcribing using SignWriting (SW). SW is devised to allow deaf people and linguists alike to exploit an easy-to-grasp written form of (any) sign language. Similarly, SWift has been developed for everyone who masters SW, and is not exclusively deaf-oriented. Using SWift, it is possible to compose and save any sign, using elementary components called glyphs. A guided procedure facilitates the composition process. SWift is aimed at helping to break down the "electronic" barriers that keep the deaf community away from Information and Communication Technology (ICT). The editor has been developed modularly and can be integrated everywhere the use of SW, as an alternative to written vocal language, may be advisable.

This paper describes the design and development of a location-based Augmented Reality (AR) application for mobile devices. The application provides real time data about transportation in a urban area. It can be set along the line of the continuous creation of richer and more complex interaction modalities between users and data. A relevant element in this strategy is the visual enrichment of the real scene perceived though the mobile camera, by superimposing to it a set of user-relevant information. In the presented work, this information is related to nearby bus stops and to the arrival of next buses. More details, such as routes, distances etc. can be displayed on demand in order to gain awareness of the surrounding infomobility data. The presented application is included in a multiservice framework named RMob, developed for the city of Rome.

In the proposed demo, a new approach to e-learning environments for deaf people is illustrated. Using a fully iconic web-based environment, a tutor can define, generate and test e-learning courses for deaf people, which are automatically managed, published and served by the system itself.

SWift (SignWriting improved fast transcriber) is an advanced editor for SignWriting (SW). At present, SW is a promising alternative to provide documents in an easyto- grasp written form of (any) Sign Language, the gestural way of communication which is widely adopted by the deaf community. SWift was developed for SW users, either deaf or not, to support collaboration and exchange of ideas. The application allows composing and saving the desired signs using elementary components called glyphs. The procedure that was devised guides and simplifies the editing process. SWift aims at breaking the "electronic" barriers that keep the deaf community away from ICT in general, and from e-learning in particular. The editor can be contained in a pluggable module; therefore, it can be integrated everywhere the use of SW might is an advisable alternative to written "verbal" language, which often hinders information grasping by deaf users.

We investigate the integration of LECOMPS, a web-based e-learning environment for the automated construction and adaptive delivery of learning paths, and SOCIALX, a web-based system for shared e-learning activities, which exploits a reputation system to provide feedback to its participants. Our overall goal is the integration of personalized and collaborative learning to support the Vygotskij's educational theory of proximal development. Therefore we propose a two-way tunneling strategy: the LECOMPS student model is used to select the set of social activities (met in SOCIALX) according to the present individual learner state of knowledge, on the other hand, the solution of exercises, and the associated reputation derived in SOCIALX, is used to update the LECOMPS student model. In particular, we present a mapping between the student model and the definition of Vygotskij's concepts of Autonomous Problem Solving and Proximal Development regions, with the aim to provide the learner with better guidance during the taking of the course.

VIVIE is a system for video sequence indexing. Video frames are annotated according to the identities of appearing subjects. Different interacting modules perform different processing steps, and each can be possibly substituted with a different one performing the same task using a different method. Classification and clustering are the most challenging activities. Differently from most existing similar systems, VIVIE accounts for the concomitant appearance of two identities in the same clip, and exploits such information for identity mapping. VIVIE was tested on 7 video clips and on a subset of the SCFace database to assess its performances.

Research on biometrics has noticeably increased. However, no single bodily or behavioral feature is able to satisfy acceptability, speed, and reliability constraints of authentication in real applications. The present trend is therefore toward multimodal systems. In this paper, we deal with some core issues related to the design of these systems and propose a novel modular framework, namely, novel approaches for biometric systems (NABS) that we have implemented to address them. NABS proposal encompasses two possible architectures based on the comparative speeds of the involved biometries. It also provides a novel solution for the data normalization problem, with the new quasi-linear sigmoid (QLS) normalization function. This function can overcome a number of common limitations, according to the presented experimental comparisons. A further contribution is the system response reliability (SRR) index to measure response confidence. Its theoretical definition allows to take into account the gallery composition at hand in assigning a system reliability measure on a single-response basis. The unified experimental setting aims at evaluating such aspects both separately and together, using face, ear, and fingerprint as test biometries. The results provide a positive feedback for the overall theoretical framework developed herein. Since NABS is designed to allow both a flexible choice of the adopted architecture, and a variable compositions and/or substitution of its optional modules, i.e., QLS and SRR, it can support different operational settings.

Virtual simulators have been used in the last twenty years for applications ranging from flight simulation to computer-based training, just to name a few. More recently a new level of simulation has been introduced thanks to haptic interfaces able to reproduce kinesthetic and/or tactile feedback, typically experimented during interaction with real-world objects. In this paper visual simulation and haptic interfaces are integrated in a novel training system for Intra Cytoplasmic Sperm Injection (ICSI), an in-vitro fertilization technique which is now a standard for the treatment of human infertility. We describe a virtual micromanipulation simulator made by two hand-based Cyberforce haptic devices (a synthetic replica of the actual manipulation gear) and a visual-haptic engine simulating the shape and the dynamic behavior of the main components in the artificial fertilization process: the human egg, the selected sperm and the micro needles required to inject the latter into the egg's cytoplasm. Our first tests, conducted so far, are encouraging.

Though face recognition gained significant attention and credibility in the last decade, quite few commercial applications were able to benefit from this. In this paper we propose FACE (Face analysis for Commercial Entities), a robust framework to address face analysis, aiming at supporting the activities of various Commercial Entities. In particular, we present two case studies, with the related experimental results which sustain the presented approach.

Advances in processing procedures make the iris a realistic candidate to the role of biometry of the future. Precise detection and segmentation for such biometry are a crucial ongoing research area. We propose an iris segmentation technique and show that it is more reliable than existent ones.

Face recognition is widely considered as one of the most promising biometric techniques, allowing high recognition rates without being too intrusive. Many approaches have been presented to solve this special pattern recognition problem, also addressing the challenging cases of face changes, mainly occurring in expression, illumination, or pose. On the other hand, less work can be found in literature that deals with partial occlusions (i.e., sunglasses and scarves). This paper presents face recognition against occlusions and expression variations (FARO) as a new method based on partitioned iterated function systems (PIFSs), which is quite robust with respect to expression changes and partial occlusions. In general, algorithms based on PIFSs compute a map of self-similarities inside the whole input image, searching for correspondences among small square regions. However, traditional algorithms of this kind suffer from local distortions such as occlusions. To overcome such limitation, information extracted by PIFS is made local by working independently on each face component (eyes, nose, and mouth). Distortions introduced by likely occlusions or expression changes are further reduced by means of anad hocdistance measure. In order to experimentally confirm the robustness of the proposed method to both lighting and expression variations, as well as to occlusions, FARO has been tested using AR-faces database, one of the main benchmarks for the scientific community in this context. A further validation of FARO performances is provided by the experimental results produced on Face Recognition Grand Challenge database.

We present a mobile game to play a museum treasure hunt, addressed to students that are about 11-14. They have to search for the " materializations" of the solutions to a sequence of riddles, and to photograph them by personal camera phones. The letters of a secret word are orderly provided on right answers, spurring the interest for the exhibition through the cellular phone. The novelty is the use of QR-Codes, a kind of 2D codes, to identify the correct answers and to enjoy some other services. A preliminary field test in the Norsk Telemuseum gave very good results. Copyright 2009 ACM.

The structure of a fractal based image indexing system is described. FINE implies image space linearization, a custom clustering strategy, Peano-serialized spatial addressing, ad-hoc heuristics for improving search and specially defined distance functions. The resulting system is invariant, or at least robust, to a large class of typical variations that appear in natural images including rotations, scaling, and changes in color or illumination. The performance of FINE is illustrated, discussed and compared with other present alternatives using standard and custom-based image databases.

We present AXEL, a WYSIWYG editor for LMML (Learning Material Markup Language). LMML augments educational content with meta-information about its components, according to Passau Teachware Model (1999). LMML is a complex markup language. AXEL (Advanced XML Editor for Learning material) allows any user to easily exploit it. We also enriched the model with scaffolding components that would otherwise require non-trivial programming abilities.

The new challenge for designers and HCI researchers is to develop software tools for effective e-learning. Learner-Centered Design (LCD) provides guidelines to make new learning domains accessible in an educationally productive manner. A number of new issues have been raised because of the new "vehicle" for education. Effective e-learning systems should include sophisticated and advanced functions, yet their interface should hide their complexity, providing an easy and flexible interaction suited to catch students' interest. In particular, personalization and integration of learning paths and communication media should be provided.It is first necessary to dwell upon the difference between attributes for platforms (containers) and for educational modules provided by a platform (contents). In both cases, it is hard to go deeply into pedagogical issues of the provided knowledge content. This work is a first step towards identifying specific usability attributes for e-learning systems, capturing the peculiar features of this kind of applications. We report about a preliminary users study involving a group of e-students, observed during their interaction with an e-learning system in a real situation. We then propose to adapt to the e-learning domain the so called SUE (Systematic Usability Evaluation) inspection, providing evaluation patterns able to drive inspectors' activities in the evaluation of an e-learning tool.

We present a system supporting synchronous distance multimedia collaboration. Such systems require a number of tools to communicate, enforce collaboration awareness and maintain consistency of shared objects. We introduce the main requirements to provide virtual room facilities, and present our implementation comparing it to other systems