Dementia represents a global public health concern, with the early detection of Alzheimer's disease, the most prevalent form of dementia, being of paramount importance. Given the limited availability of suitable biomarkers, research has shown that early cognitive impairment can be identified through patients' spoken language. This paper presents a multi-modal system for automatic Alzheimer's disease detection using speech. The system has been trained on spoken recordings of healthy individuals and Alzheimer's patients describing an image, a task requiring linguistic and cognitive skills. Built on fine-tuned advanced Large Language Models, audio feature extractors, and classifiers, the system, after an extensive comparison of single and multi-modal architectures, achieves optimal results with the combination of Mistral-7B, VGGish, and Support Vector Classifier, outperforming previous methods on the ADReSSo 2021 test set.

The widespread adoption of masks and respirators has significantly influenced various aspects of society, driving technological advances to improve comfort, efficiency, and sustainability. The COVID-19 pandemic underscored their essential role in the protection of public health, with continued relevance in the industrial, environmental, and hygiene-critical sectors. Recent developments in deep learning offer promising approaches for building automated systems that can detect mask and respirator usage. In this regard, this paper first aims to present an improved version of the Facial Masks and Respirators Database (FMR-DB), which can be used to create such systems. New features include a significant increase in available images, which has been expanded from 2565 to 4200 images, and the addition of You Only Look Once (YOLO), PASCAL Visual Object Classes (PascalVOC), and Common Objects in Context (COCO) labeling for image detection tasks. Furthermore, image classification and object detection tests were conducted using Convolutional Neural Networks (CNNs), Transformers, and YOLO to determine the types of masks and respirators accurately. Finally, to the best of the authors' knowledge, these tools were used for the first time to analyze the protection levels of respirators automatically. The results provide valuable insights for developing efficient and reliable automatic recognition systems.

Multidimensional 2D-3D face analysis has demonstrated a strong potential for human identification in several application domains. The combined, synergic use of 2D and 3D data from human faces can counteract typical limitations in 2D face recognition, while improving both accuracy and robustness in identification. On the other hand, current mobile devices, often equipped with depth cameras and high performance computing resources, offer a powerful and practical tool to better investigate new models to jointly process real 2D and 3D face data. However, recent concerns related to privacy of individuals and the collection, storage and processing of personally identifiable biometric information have diminished the availability of public face recognition datasets. Uniss-MDF (Uniss-MultiDimensional Face) represents the first collection of combined 2D-3D data of human faces captured with a mobile device. Over 76,000 depth images and videos are captured from over 100 subjects, in both controlled and uncontrolled conditions, over two sessions. The features of Uniss-MDF are extensively compared with existing 2D-3D face datasets. The reported statistics underscore the value of the dataset as a versatile resource for researchers in face recognition on the move and for a wide range of applications. Notably, it is the sole 2D-3D facial dataset using data from a mobile device that includes both 2D and 3D synchronized sequences acquired in controlled and uncontrolled conditions. The Uniss-MDF dataset and the proposed experimental protocols with baseline results provide a new platform to compare processing models for novel research avenues in advanced face analysis on the move.

This research focuses on developing improved diagnostic tools for Alzheimer's Disease (AD), a condition impacting approximately 50 million individuals globally. In the paper, we achieve automatic AD detection by leveraging pre-trained Large Language Models (LLMs) for linguistic analysis applied to the ADReSS/ADReSSo (Alzheimer's Dementia Recognition through Spontaneous Speech/only) Challenges datasets, following speech-to-text conversion. While the recent advancements in LLMs offer a robust foundation for their application in healthcare, fine-tuning these models for specific tasks, like AD detection, requires optimization to balance performance and computational efficiency. Also in response to data privacy concerns in healthcare, we implement our methodology on consumer-level GPU cards, which offer a practical solution for local data processing. Our approach uses fine-tuning techniques such as Low Ranking Adaptation and Parameter-Efficient Fine-Tuning to enhance the capabilities of Large Language Models within the limits of consumer-grade hardware. Additionally, we incorporate quantization to reduce computational demands while preserving model accuracy. Conducted on setups with RTX 4090 and dual RTX 3090 GPUs, our experiments demonstrate promising results that align with or surpass existing benchmarks in dementia recognition.

In recent years the number of credentials and digital identities needed to access online services is exploding, increasing the risk of fraud and unauthorized access to personal data.Biometric verification techniques can help to build more secure authentication systems; in particular face recognition algorithms have reached near perfect accuracy under ideal conditions. In Europe 60% of population owns an electronic identification (eID) document which contains biometric data, and almost all member states provide tools to verify the identity of a user who holds an eID card; however, the adoption of biometric authentication schemes is still limited. Reasons for this low adoption are privacy concerns, about the storage or transmission of personal images, and the degradation of performance of facial recognition algorithms with 2D images acquired "in the wild" (real world condition with inconsistency in lighting and positioning, unclear or occluded facial features, noise from the background). In this paper a new authentication scheme is proposed, exploiting images acquired "in the wild" and 3D facial reconstruction techniques which can help in improving the quality of the probed images, as changes in lighting or position have limited impact on facial shape. Several experiments have been conducted in order to investigate the performance of the proposed method, using the Bosphorus and the NIST Face Recognition Grand Challenge (FRGC) datasets. In particular, one of the most accurate ML face recognition algorithms has been selected to compare the performance of 2D snapshot data, obtained from 3D back-projection, against 2D uncontrolled images acquired "in the wild". We show that using 3D facial reconstructions significantly improve face recognition performance, allowing for a more secure and robust authentication system.

Abstract Traditional local image descriptors such as SIFT and SURF are based on processings similar to those that take place in the early visual cortex. Nowadays, convolutional neural networks still draw inspiration from the human vision system, integrating computational elements typical of higher visual cortical areas. Deep CNN’s architectures are intrinsically hard to interpret, so much effort has been made to dissect them in order to understand which type of features they learn. However, considering the resemblance to the human vision system, no enough attention has been devoted to understand if the image features learned by deep CNNs and used for classification correlate with features that humans select when viewing images, the so-called human fixations, nor if they correlate with earlier developed handcrafted features such as SIFT and SURF. Exploring these correlations is highly meaningful since what we require from CNNs, and features in general, is to recognize and correctly classify objects or subjects relevant to humans. In this paper, we establish the correlation between three families of image interest points: human fixations, handcrafted and CNN features. We extract features from the feature maps of selected layers of several deep CNN’s architectures, from the shallowest to the deepest. All features and fixations are then compared with two types of measures, global and local, which unveil the degree of similarity of the areas of interest of the three families. From the experiments carried out on ETD human fixations database, it turns out that human fixations are positively correlated with handcrafted features and even more with deep layers of CNNs and that handcrafted features highly correlate between themselves as some CNNs do.

Early hierarchical computational visual models as well as recent deep neural networks have been inspired by the functioning of the primate visual cortex system. Although much effort has been made to dissect neural networks to visualize the features they learn at the individual units, the scope of the visualizations has been limited to a categorization of the features in terms of their semantic level. Considering the ability humans have to select high semantic level regions of a scene, the question whether neural networks can match this ability, and if similarity with humans attention is correlated with neural networks performance naturally arise. To address this question we propose a pipeline to select and compare sets of feature points that maximally activate individual networks units to human fixations. We extract features from a variety of neural networks, from early hierarchical models such as HMAX up to recent deep convolutional neural netwoks such as Densnet, to compare them to human fixations. Experiments over the ETD database show that human fixations correlate with CNNs features from deep layers significantly better than with random sets of points, while they do not with features extracted from the first layers of CNNs, nor with the HMAX features, which seem to have low semantic level compared with the features that respond to the automatically learned filters of CNNs. It also turns out that there is a correlation between CNN's human similarity and classification performance.

The adoption of Cloud computing involves many advantages in terms of flexibility, scalability and reliability, but also implies new challenges on security, data privacy and protection of personal data. Since more and more sensitive applications and data are moved to the cloud, the verification of the digital identity of the participants in the electronic communication has become a crucial challenge. Currently, the use of biometric techniques can be considered as an effective solution to ensure a significant increase of security in the authentication protocols managed by modern authentication servers. However the use of biometric data for the logical access to IT services is a more challenging and still unsolved problem. The project Cloud for SME integrates a biometric authentication based on fingerprints with a cloud computing platform, investigating how highly secure authentication methods can increase the adoption of cloud computing technologies among small and medium enterprises.

Modelling behavioural biometric patterns is a key issue for modern user centric applications, aimed at better monitoring users' activities, understanding their habits and detecting their identity. Following this trend, this paper investigates whether the electrical energy consumption of a user can be a distinctive behavioural biometric trait. In particular we analyse daily and weekly load profiles showing that they are closely related to the identity of the users. Hence, we believe that this level of analysis can open interesting application scenarios in the field of energy management and it provides a good working framework for the continuous development of smart environments with demonstrable benefits on real-world implementations.

When dealing with face recognition, multimodal algorithms, with their potential to capture complementary characteristics from the 2D and 3D data channels, can reach high level of efficiency and robustness. In this paper, we explore different combinations of iconic descriptors coupled with a shape descriptor and propose a fully automatic, multimodal, face recognition paradigm. Two iconic features extractors, the Scale Invariant Feature Transform (SIFT) and the Speeded-Up Robust Features (SURF), are used, in turn, to extract salient points from the images of the faces. The corresponding points on the scans are validated with Joint Differential Invariants, a 3D characterisation method based on local and global shape information. SIFT and SURF are then combined at feature level and the 3D Joint Differential Invariants used to validate them on the shape channel. The proposed method has been tested on the FRGCv2 database. Experimental results highlight the complementarity of the feature points extracted by SIFT and SURF and the effectiveness of their 3D validation.

The paper proposes an original method, derived from basic face recognition and classification research, which is a good candidate for an effective automotive application. The proposed approach exploits a single b/w camera, positioned in front of the driver, and a very efficient classification strategy, based on neural network classifiers. A peculiar feature of the work is the adoption of iconic data reduction, avoiding specific and time-consuming feature-based approaches. Though at an initial development stage, the method proved to be fast and robust compared to state of the art techniques; experimental results show real-time response and mean weighted accuracy near to 93%. The method requires a simple training procedure which can be certainly improved for real applications; moreover it can be easily integrated with techniques for automatic face-recognition of the driver.

Visual surveillance in outdoor environments requires the monitoring of both objects and events. The analysis is generally driven by the target application which, in turn, determines the set of relevant events and objects to be analyzed. In this paper we concentrate on the analysis of outdoor scenes, in particular for vehicle traffic control. In this scenario, the analysis of weather conditions is considered to signal particular and potentially dangerous situations like the presence of snow, fog, or heavy rain. The developed system uses a statistical framework based on the mixture of Gaussians to identify changes both in the spatial and temporal frequencies which characterize specific meteorological events. Several experiments performed on standard databases and real scenes demonstrate the applicability of the proposed approach.

It has been shown in many different contexts that the Generalized Gaussian (GG) distribution represents a flexible and suitable tool for data modeling. Almost all the reported applications are focused on modeling points (fixed length vectors); a different but crucial scenario, where the employment of the GG has received little attention, is the modeling of sequential data, i.e. variable length vectors. This paper explores this last direction, describing a variant of the well known Hidden Markov Model (HMM) where the emission probability function of each state is represented by a GG. A training strategy based on the Expectation Maximization (EM) algorithm is presented. Different experiments using both synthetic and real data (EEG signal classification and face recognition) show the suitability of the proposed approach compared with the standard Gaussian HMM.

This paper develops and demonstrates an original approach to face-image analysis based on identifying distinctive areas of each individual's face by its comparison to others in the population. The method differs from most others—that we refer as unary —where salient regions are defined by analyzing only images of the same individual. We extract a set of multiscale patches from each face image before projecting them into a common feature space. The degree of “distinctiveness” of any patch depends on its distance in feature space from patches mapped from other individuals. First a pairwise analysis is developed and then a simple generalization to the multiple-face case is proposed. A perceptual experiment, involving 45 observers, indicates the method to be fairly compatible with how humans mark faces as distinct. A quantitative example of face authentication is also performed in order to show the essential role played by the distinctive information. A comparative analysis shows that performance of our n-ary approach is as good as several contemporary unary, or binary, methods, while tapping a complementary source of information. Furthermore, we show it can also provide a useful degree of illumination invariance.

As confirmed by recent neurophysiological studies, the use of dynamic information is extremely important for humans in visual perception of biological forms and motion. Apart from the mere computation of the visual motion of the viewed objects, the motion itself conveys far more information, which helps understanding the scene. This paper provides an overview and some new insights on the use of dynamic visual information for face recognition. In this context, not only physical features emerge in the face representation, but also behavioral features should be accounted. While physical features are obtained from the subject's face appearance, behavioral features are obtained from the individual motion and articulation of the face. In order to capture both the face appearance and the face dynamics, a dynamical face model based on a combination of hidden Markov models is presented. The number of states (or facial expressions) are automatically determined from the data by unsupervised clustering of expressions of faces in the video. The underlying architecture closely recalls the neural patterns activated in the perception of moving faces. Preliminary results on real video image data show the feasibility of the proposed approach

The paper presents a technical solution for the automation of network inventory management (NIM) processes for mobile operators. The deployment of operations support systems (OSS), according to the design of an overall target systems architecture, and the reengineering of management processes supporting NIM are the core issues of the proposed solution. Common information and data models are key enabling factors for NIM processes and systems. Information modelling is also the core of standardization efforts for network and service management. Within the operating environments of telecommunication companies (telcos), running either wireline or wireless networks, NIM and the centrality of data have appeared to be key factors for enabling automation of network and service management and deploying effective OSS solutions. The theories and the results discussed have been demonstrated practically and successfully in the operating companies of the Telecom Italia Group. Software solutions for the automation of NIM processes have been deployed since the early 1980s, using early software technologies for database management and building bespoke systems, until some more recent implementations that exploit best-in-class commercial products and successful Web technologies. A practical case study is presented of a successful implementation of an OSS solution for NIM in a mobile company of the Telecom Italia Group.

The use of biometric data for user authentication and/or recognition is now a reality. On the other hand, there is still a strong need for new technologies to overpass intrinsic limitations of already "established" techniques. This not only requires to devise new algorithms but to determine the real potential and limitations of existing techniques. This is possible only devising standard testing and assessment procedures based on statistical observations of the outputs of the system. In order to define better standard evaluation process, a system based on space-variant iconic image matching is described and the validation procedure defined. It turns out that all methods based on the same biometric measurements have the same intrinsic limitations, which can be only overcome by the adoption of a multi-modal or multi-algorithmic approach.

Algorithms based on face images are quite appealing for the possibility to easily adapt and tailor a system to many application domains. A system for personal identity verification and also recognition is presented. The core engine is a standard correlation-based matcher performed on iconic representations of face images. Through standard statistical tests of the recognition results obtained from two different data sets the actual physical limits of the pattern matcher are clearly shown. Successively also other aspects are taken into account, related to the feature space, allowing us to greatly improve the system performance reaching almost 100% correct recognition.

This paper describes a novel approach to the problem of reaching an object in space under visual guidance. The approach is characterized by a great robustness to calibration errors, such that virtually no calibration is required. Servoing is based on binocular vision: a continuous measure of the end-effector motion field, derived from real-time computation of the binocular optical flow over the stereo images, is compared with the actual position of the target and the relative error in the end-effector trajectory is continuously corrected. The paper outlines the general framework of the approach, shows how visual measures are obtained and discusses the synthesis of the controller along with its stability analysis. Real-time experiments are presented to show the applicability of the approach in real 3-D applications.

Visual navigation is a challenging issue in automated robot control. In many robot applications, like object manipulation in hazardous environments or autonomous locomotion, it is necessary to automatically detect and avoid obstacles while planning a safe trajectory. In this context the detection of corridors of free space along the robot trajectory is a very important capability which requires nontrivial visual processing. In most cases it is possible to take advantage of the active control of the cameras. In this paper we propose a cooperative schema in which motion and stereo vision are used to infer scene structure and determine free space areas. Binocular disparity, computed on several stereo images over time, is combined with optical flow from the same sequence to obtain a relative-depth map of the scene. Both the time to impact and depth scaled by the distance of the camera from the fixation point in space are considered as good, relative measurements which are based on the viewer, but centered on the environment. The need for calibrated parameters is considerably reduced by using an active control strategy. The cameras track a point in space independently of the robot motion and the full rotation of the head, which includes the unknown robot motion, is derived from binocular image data. The feasibility of the approach in real robotic applications is demonstrated by several experiments performed on real image data acquired from an autonomous vehicle and a prototype camera head.>

This paper deals with the problem of the vision-based feedback control of an eye-head system. Movements of the visual sensors are controlled in order to gather optimal measurements, in relation to the kinematic structure of the system and the task that is currently being executed. In more explicit biological terms, it is the problem of, given a particular sensor, how to use movement to ameliorate perception. In particular it is proved that movements can essentially be guided by the achievement of the minimum cost function related with the sensitivity of the transformation from world to camera coordinates. This allows one to define proper control objectives, for the coordinated eye-head movements; it also formally motivates some relevant aspects of the biological vision like fixation, vergence and eye-head compensation.>

A general framework for dynamic stereo is proposed. It is shown that a binocular system, able to actively control the gaze of the cameras, can profitably exploit stereo and motion analysis to compute the relative orientation and time-to-impact with respect to the environment. The cooperative schema, in which motion and stereo vision are combined to take under control the temporal evolution of disparity, is particularly suitable for robot navigation or for tasks involving a robotic head mounted on a mobile platform. The framework takes into account tilt and pan motion of the cameras, and generic translation of the binocular system. An experiment performed on a stereo image sequence from a prototype robotic head is presented.>

The authors consider orienting, that is, establishing and maintaining a spatial relation between a motorized pair of cameras (the eye-head system) and a static or a moving object tracked over time. Motivated by physiological evidence, they propose a simple set of vision-based strategies aimed to perform head, eye, and body movements in a complex environment. Fixation is shown to be an essential feature in visual servoing, and is used to decouple control on head rotational degrees of freedom, making possible a metric-less approach to the orientation problem. An implementation of these strategies, using a binocular camera system mounted on a PUMA 700 robotic system, demontrated the effectiveness of the approach.>

The paper describes the practical implementation of a vision-based navigation system for a mobile robot operating in indoor environments. The robot acquires visual information by means of three CCD cameras mounted on board. A stereo pair is used for ground plane obstacle detection and avoidance, while the third camera is used to locate landmarks and compute the robot's position. Odometric readings are used to guide visual perception by simple 'where to look next' strategies. The whole processing and the control architecture determining the overall behaviour of the robot are mainly implemented on a parallel MIMD machine. Some examples are presented, showing how the robot moves in a partially structured environment reaching the specified goal points with a fair degree of accuracy, avoiding unpredicted obstacles and following trajectories obtained through the cooperation of the various navigation modules running in parallel.>

A cooperative schema is proposed in which binocular disparity, computed on several stereo images over time, is combined with optical flow from the same sequence to obtain a relative-depth map of the scene. Both time-to-impact and depth scaled by the distance of the camera from the fixation point in space are considered as good, relative measurements which are based on the viewer (but centered on the environment). Two experiments, performed on image sequences from real scenes, are presented.>

The authors describes an approach to the visuo-motor coordination of the grasping problem. A vision system is used to provide a spatial description of the scene. The volumetric information is reduced by merging the voxels into an octree representation. At the end of the reconstruction process, this kind of 'essential' information can be used to choose the right pre-shaping and to coordinate the arm and the hand motion towards the pre-shaping position, in order to realize a stable grasp. Some preliminary results are shown.>

The extraction of reliable range data from images is investigated, considering, as a possible solution, the integration of different sensor modalities. Two different algorithms are used to obtain independent estimates of depth from a sequence of stereo images. The results are integrated on the basis of the uncertainty of each measure. The stereo algorithm uses a coarse-to-fine control strategy to compute disparity. An algorithm for depth-from-motion is used, exploiting the constraint imposed by active motion of the cameras. To obtain a 3D description of the objects, the motion of the cameras is purposefully controlled, in such a manner as to move around the objects in view while the gaze is directed toward a fixed point in space. This egomotion strategy, which is similar to that adopted by the human visuomotor system, allows a better exploration of partially occluded objects and simplifies the motion equations. When tested on real scenes, the algorithm demonstrated a low sensitivity to image noise, mainly due to the integration of independent measures. An experiment performed on a real scene containing several objects is presented.>