Search

Highlights • We introduce trainable COPE feature extractors for sound representation learning. • A COPE feature extractor is trained using a single prototype sound of interest. • We propose a method for audio event detection with robustness to varying SNR. • We experiment on four data sets: MIVIA audio, MIVIA roads, ESC-10 and TU Dortmund. • We achieve better results than existing methods for audio event detection. Abstract Sound analysis research has mainly been focused on speech and music processing. The deployed methodologies are not suitable for analysis of sounds with varying background noise, in many cases with very low signal-to-noise ratio (SNR). In this paper, we present a method for the detection of patterns of interest in audio signals. We propose novel trainable feature extractors, which we call COPE (Combination of Peaks of Energy). The structure of a COPE feature extractor is determined using a single prototype sound pattern in an automatic configuration process, which is a type of representation learning. We construct a set of COPE feature extractors, configured on a number of training patterns. Then we take their responses to build feature vectors that we use in combination with a classifier to detect and classify patterns of interest in audio signals. We carried out experiments on four public data sets: MIVIA audio events, MIVIA road events, ESC-10 and TU Dortmund data sets. The results that we achieved (recognition rate equal to 91.71% on the MIVIA audio events, 94% on the MIVIA road events, 81.25% on the ESC-10 and 94.27% on the TU Dortmund) demonstrate the effectiveness of the proposed method and are higher than the ones obtained by other existing approaches. The COPE feature extractors have high robustness to variations of SNR. Real-time performance is achieved even when the value of a large number of features is computed. [ABSTRACT FROM AUTHOR]

This paper is a historical overview of graph-based methodologies in Pattern Recognition in the last 40 years; history is interpreted with the aim of recognizing the rationale inspiring the papers published in these years, so as to roughly classify them. Despite the extent of scientific production in this field, it is possible to identify three historical periods, each having its own connotation common to most of the corresponding papers, which are called here as the pure, the impure and extreme periods. [ABSTRACT FROM AUTHOR]

Age estimation from face images can be profitably employed in several applications, ranging from digital signage to social robotics, from business intelligence to access control. Only in recent years, the advent of deep learning allowed for the design of extremely accurate methods based on convolutional neural networks (CNNs) that achieve a remarkable performance in various face analysis tasks. However, these networks are not always applicable in real scenarios, due to both time and resource constraints that the most accurate approaches often do not meet. Moreover, in case of age estimation, there is the lack of a large and reliably annotated dataset for training deep neural networks. Within this context, we propose in this paper an effective training procedure of CNNs for age estimation based on knowledge distillation, able to allow smaller and simpler "student" models to be trained to match the predictions of a larger "teacher" model. We experimentally show that such student models are able to almost reach the performance of the teacher, obtaining high accuracy over the LFW+, LAP 2016 and Adience datasets, but being up to 15 times faster. Furthermore, we evaluate the performance of the student models in the presence of image corruptions, and we demonstrate that some of them are even more resilient to these corruptions than the teacher model. [ABSTRACT FROM AUTHOR]

Although in recent years we have witnessed an explosion of the scientific research in the recognition of facial soft biometrics such as gender, age and expression with deep neural networks, the recognition of ethnicity has not received the same attention from the scientific community. The growth of this field is hindered by two related factors: on the one hand, the absence of a dataset sufficiently large and representative does not allow an effective training of convolutional neural networks for the recognition of ethnicity; on the other hand, the collection of new ethnicity datasets is far from simple and must be carried out manually by humans trained to recognize the basic ethnicity groups using the somatic facial features. To fill this gap in the facial soft biometrics analysis, we propose the VGGFace2 Mivia Ethnicity Recognition (VMER) dataset, composed by more than 3,000,000 face images annotated with 4 ethnicity categories, namely African American, East Asian, Caucasian Latin and Asian Indian. The final annotations are obtained with a protocol which requires the opinion of three people belonging to different ethnicities, in order to avoid the bias introduced by the well-known other race effect. In addition, we carry out a comprehensive performance analysis of popular deep network architectures, namely VGG-16, VGG-Face, ResNet-50 and MobileNet v2. Finally, we perform a cross-dataset evaluation to demonstrate that the deep network architectures trained with VMER generalize on different test sets better than the same models trained on the largest ethnicity dataset available so far. The ethnicity labels of the VMER dataset and the code used for the experiments are available upon request at . [ABSTRACT FROM AUTHOR]

• Analysis of recent trends of graph-based techniques for Pattern Recognition. • Based on the 11th IAPR-TC15 Workshop on Graph-based Representations for PR (GbR2017). • Most popular techniques are metrics on graphs and inexact graph matching. In this paper we try to examine recent trends on the use of graph-based representations in Pattern Recognition, using as a vantage point the 11th IAPR-TC15 Workshop GbR2017, dedicated to this topic. A survey of the paper presented at GbR2017 will give us the opportunity to reflect on the directions where the interest of the research community working on this subject is moving. [ABSTRACT FROM AUTHOR]

In industrial environments, it is crucial to establish a strong collaboration between humans and robots to enhance productivity. However, the nature of the work demands that workers have the authority to provide specific instructions to the robots. The scientific community has extensively investigated these dual requirements, aiming to develop advanced systems capable of recognizing voice commands and implementing speaker authentication. Nevertheless, in the industrial context, these tasks should be executed simultaneously on low-cost and low-power embedded devices that can be mounted on board the robotic platform. To overcome this challenge, we propose a multi-task network for Speech-Command Recognition and Speaker Identification. Additionally, we employ the GradNorm adaptive algorithm to address the issue of task imbalance. To evaluate the proposed system, we introduce a new dataset, MIVIA-ISC, consisting of 20,857 samples uttered by 562 speakers for 31 distinct commands. Our approach significantly reduces the network size by 47% and its execution time by 48% compared to the commonly used methodology, which employs one network for each task. Furthermore, our approach demonstrates a significant improvement in the accuracy of the Speaker Identification task, achieving an 11% increase compared to the corresponding single-task network. Importantly, this enhancement is achieved without compromising the accuracy of the Speech-Command Recognition task, which experiences only a minimal 3% decrease in performance. • We propose a multi-task approach for speech and speaker recognition. • A speech command and speaker recognition dataset in an industrial context. • The method reduces the computational load and increases the robustness of the system. • The system runs in real-time on an embedded device. [ABSTRACT FROM AUTHOR]

Highlights • We propose trainable feature extractors for representation of skeleton poses. • We employ the proposed feature extractor for classification of human actions. • We designed an human action classification method based on string kernel. • We carried out experiments on the MHAD, MSRDA and MIVIA-S action datasets. • We publicly released the MIVIA-S dataset for research purpose. Abstract Automatic interpretation of human actions gained strong interest among researchers in patter recognition and computer vision because of its wide range of applications, such as in social and home robotics, elderly people health care, surveillance, among others. In this paper, we propose a method for recognition of human actions by analysis of skeleton poses. The method that we propose is based on novel trainable feature extractors, which can learn the representation of prototype skeleton examples and can be employed to recognize skeleton poses of interest. We combine the proposed feature extractors with an approach for classification of pose sequences based on string kernels. We carried out experiments on three benchmark data sets (MIVIA-S, MSRSDA and MHAD) and the results that we achieved are comparable or higher than the ones obtained by other existing methods. A further important contribution of this work is the MIVIA-S dataset, that we collected and made publicly available. [ABSTRACT FROM AUTHOR]

Current state-of-the-art audio analysis algorithms based on deep learning rely on hand-crafted Spectrogram-like audio representations, that are more compact than descriptors obtained from the raw waveform; the latter are, in turn, far from achieving good generalization capabilities when few data are available for the training. However, Spectrogram-like representations have two main limitations: (1) The parameters of the filters are defined a priori, regardless of the specific audio analysis task; (2) such representations do not perform any denoising operation on the audio signal, neither in the time domain nor in the frequency domain. To overcome these limitations, we propose a new general-purpose convolutional architecture for audio analysis tasks that we call DEGramNet, which is trained with audio samples described with a novel, compact and learnable time–frequency representation that we call DEGram. The proposed representation is fully trainable: Indeed, it is able to learn the frequencies of interest for the specific audio analysis task; in addition, it performs denoising through a custom time–frequency attention module, which amplifies the frequency and time components in which the sound is actually located. It implies that the proposed representation can be easily adapted to the specific problem at hands, for instance giving more importance to the voice frequencies when the network needs to be used for speaker recognition. DEGramNet achieved state-of-the-art performance on the VGGSound dataset (for Sound Event Classification) and comparable accuracy with a complex and special-purpose approach based on network architecture search over the VoxCeleb dataset (for Speaker Identification). Moreover, we demonstrate that DEGram allows to achieve high accuracy with lightweight neural networks that can be used in real-time on embedded systems, making the solution suitable for Cognitive Robotics applications. [ABSTRACT FROM AUTHOR]

Hand washing preparation can be considered as one of the main strategies for reducing the risk of surgical site contamination and thus the infections risks. Within this context, in this paper we propose an embedded system able to automatically analyze, in real-time, the sequence of images acquired by a depth camera to evaluate the quality of the handwashing procedure. In particular, the designed system runs on an NVIDIA Jetson Nano TM computing platform. We adopt a convolutional neural network, followed by a majority voting scheme, to classify the movement of the worker according to one of the ten gestures defined by the World Health Organization. To test the proposed system, we collect a dataset built by 74 different video sequences. The results achieved on this dataset confirm the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Self-driving vehicles must be controlled by navigation algorithms that ensure safe driving for passengers, pedestrians and other vehicle drivers. One of the key factors to achieve this goal is the availability of effective multi-object detection and tracking algorithms, which allow to estimate position, orientation and speed of pedestrians and other vehicles on the road. The experimental analyses conducted so far have not thoroughly evaluated the effectiveness of these methods in road driving scenarios. To this aim, we propose in this paper a benchmark of modern multi-object detection and tracking methods applied to image sequences acquired by a camera installed on board the vehicle, namely, on the videos available in the BDD100K dataset. The proposed experimental framework allows to evaluate 22 different combinations of multi-object detection and tracking methods using metrics that highlight the positive contribution and limitations of each module of the considered algorithms. The analysis of the experimental results points out that the best method currently available is the combination of ConvNext and QDTrack, but also that the multi-object tracking methods applied on road images must be substantially improved. Thanks to our analysis, we conclude that the evaluation metrics should be extended by considering specific aspects of the autonomous driving scenarios, such as multi-class problem formulation and distance from the targets, and that the effectiveness of the methods must be evaluated by simulating the impact of the errors on driving safety. [ABSTRACT FROM AUTHOR]

The inspection of retinal fundus images allows medical doctors to diagnose various pathologies. Computer-aided diagnosis systems can be used to assist in this process. As a first step, such systems delineate the vessel tree from the background. We propose a method for the delineation of blood vessels in retinal images that is effective for vessels of different thickness. In the proposed method, we employ a set of B-COSFIRE filters selective for vessels and vessel-endings. Such a set is determined in an automatic selection process and can adapt to different applications. We compare the performance of different selection methods based upon machine learning and information theory. The results that we achieve by performing experiments on two public benchmark data sets, namely DRIVE and STARE, demonstrate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

Autonomous robots performing cooperative tasks need to know the relative pose of the other robots in the fleet. Deducing these poses might be performed through structure from motion methods in the applications where there are no landmarks or GPS, for instance, in non-explored indoor environments. Structure from motion is a technique that deduces the pose of cameras only given only the 2D images. This technique relies on a first step that obtains a correspondence between salient points of images. For this reason, the weakness of this method is that poses cannot be estimated if a proper correspondence is not obtained due to low quality of the images or images that do not share enough salient points. We propose, for the first time, an interactive structure-from-motion method to deduce the pose of 2D cameras. Autonomous robots with embedded cameras have to stop when they cannot deduce their position because the structure-from-motion method fails. In these cases, a human interacts by simply mapping a pair of points in the robots’ images. Performing this action the human imposes the correct correspondence between them. Then, the interactive structure from motion is capable of deducing the robots’ lost positions and the fleet of robots can continue their high level task. From the practical point of view, the interactive method allows the whole system to achieve more complex tasks in more complex environments since the human interaction can be seen as a recovering or a reset process. [ABSTRACT FROM AUTHOR]

In this paper, we propose a method that is able to detect fires by analyzing videos acquired by surveillance cameras. Two main novelties have been introduced. First, complementary information, based on color, shape variation, and motion analysis, is combined by a multiexpert system. The main advantage deriving from this approach lies in the fact that the overall performance of the system significantly increases with a relatively small effort made by the designer. Second, a novel descriptor based on a bag-of-words approach has been proposed for representing motion. The proposed method has been tested on a very large dataset of fire videos acquired both in real environments and from the web. The obtained results confirm a consistent reduction in the number of false positives, without paying in terms of accuracy or renouncing the possibility to run the system on embedded platforms. [ABSTRACT FROM PUBLISHER]

Retinal imaging provides a non-invasive opportunity for the diagnosis of several medical pathologies. The automatic segmentation of the vessel tree is an important pre-processing step which facilitates subsequent automatic processes that contribute to such diagnosis. We introduce a novel method for the automatic segmentation of vessel trees in retinal fundus images. We propose a filter that selectively responds to vessels and that we call B -COSFIRE with B standing for bar which is an abstraction for a vessel. It is based on the existing COSFIRE (Combination Of Shifted Filter Responses) approach. A B -COSFIRE filter achieves orientation selectivity by computing the weighted geometric mean of the output of a pool of Difference-of-Gaussians filters, whose supports are aligned in a collinear manner. It achieves rotation invariance efficiently by simple shifting operations. The proposed filter is versatile as its selectivity is determined from any given vessel-like prototype pattern in an automatic configuration process. We configure two B -COSFIRE filters, namely symmetric and asymmetric, that are selective for bars and bar-endings, respectively. We achieve vessel segmentation by summing up the responses of the two rotation-invariant B -COSFIRE filters followed by thresholding. The results that we achieve on three publicly available data sets (DRIVE: Se = 0.7655, Sp = 0.9704; STARE: Se = 0.7716, Sp = 0.9701; CHASE_DB1: Se = 0.7585, Sp = 0.9587) are higher than many of the state-of-the-art methods. The proposed segmentation approach is also very efficient with a time complexity that is significantly lower than existing methods. [ABSTRACT FROM AUTHOR]

This paper aims at dynamically understanding the properties of a scene from the analysis of moving object trajectories. Two different applications are proposed: the former is devoted to identify abnormal behaviors, while the latter allows to extract the \(k\) , most of the similar trajectories to the one hand-drawn by an human operator. A set of normal trajectories’ models is extracted using a novel unsupervised learning technique: the scene is adaptively partitioned into zones using the distribution of the training set and each trajectory is represented as a sequence of symbols by considering positional information (the zones crossed in the scene), speed, and shape. The main novelty is the use of a kernel-based approach for evaluating the similarity between the trajectories. Furthermore, we define a novel and efficient kernel-based clustering algorithm, aimed at obtaining groups of normal trajectories. Experimentations, conducted over three standard data sets, confirm the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

In this paper, we examine the main advances registered in the last ten years in Pattern Recognition methodologies based on graph matching and related techniques, analyzing more than 180 papers; the aim is to provide a systematic framework presenting the recent history and the current developments. This is made by introducing a categorization of graph-based techniques and reporting, for each class, the main contributions and the most outstanding research results. [ABSTRACT FROM AUTHOR]

Presents an algorithm for learning structural data from examples based on a formulation of the learning problem in terms of attributed relational graphs. Definition of proper inference operators as graph generalization and specialization; Symbolic learning systems; Pros and cons of symbolic and connectionist approaches.

In the last years, a big interest of both the scientific community and the market has been devoted to the design of audio surveillance systems, able to analyse the audio stream and to identify events of interest; this is particularly true in security applications, in which the audio analytics can be profitably used as an alternative to video analytics systems, but also combined with them. Within this context, in this paper we propose a novel recurrent convolutional neural network architecture, named DENet; it is based on a new layer that we call denoising-enhancement (DE) layer, which performs denoising and enhancement of the original signal by applying an attention map on the components of the band-filtered signal. Differently from state-of-the-art methodologies, DENet takes as input the lossless raw waveform and is able to automatically learn the evolution of the frequencies-of-interest over time, by combining the proposed layer with a bidirectional gated recurrent unit. Using the feedbacks coming from classifications related to consecutive frames (i.e. that belong to the same event), the proposed method is able to drastically reduce the misclassifications. We carried out experiments on the MIVIA Audio Events and MIVIA Road Events public datasets, confirming the effectiveness of our approach with respect to other state-of-the-art methodologies. [ABSTRACT FROM AUTHOR]

• Two parallelized versions of the VF3 subgraph isomorphism algorithm are shown. • The algorithms performance on several datasets is discussed. • Parallelized VF3 is efficient on large, dense graphs. • On small or sparse graphs the sequential version is more efficient. • The combination of a global and a local stack is effective in reducing overhead. An important combinatorial problem is subgraph isomorphism , which formalizes the task of searching for occurrences of a known substructure within a larger structure represented by a graph: applications are in the fields of chemistry, biology, medicine, databases, social network analysis. Subgraph isomorphism has been proven to be NP-complete in the general case, but several algorithms exist that use heuristics to achieve an affordable run time for common classes of graphs. The need of working with larger and larger graphs makes attractive the idea of parallelizing this task; however, a consensus has not yet been reached on what is the best strategy for doing so. In this paper, we present two versions of a new, parallel algorithm that is based on a re-design of the well known VF3 algorithm. We discuss the changes that were made to efficiently distribute the work among multiple processors. The algorithms have been evaluated with a comprehensive experimentation, using several publicly available graph datasets, to demonstrate their effectiveness in exploiting the parallelism. [ABSTRACT FROM AUTHOR]

Face analysis includes a variety of specific problems as face detection, person identification, gender and ethnicity recognition, just to name the most common ones; in the last two decades, significant research efforts have been devoted to the challenging task of age estimation from faces, as witnessed by the high number of published papers. The explosion of the deep learning paradigm, that is determining a spectacular increasing of the performance, is in the public eye; consequently, the number of approaches based on deep learning is impressively growing and this also happened for age estimation. The exciting results obtained have been recently surveyed on almost all the specific face analysis problems; the only exception stands for age estimation, whose last survey dates back to 2010 and does not include any deep learning based approach to the problem. This paper provides an analysis of the deep methods proposed in the last six years; these are analysed from different points of view: the network architecture together with the learning procedure, the used datasets, data preprocessing and augmentation, and the exploitation of additional data coming from gender, race and face expression. The review is completed by discussing the results obtained on public datasets, so as the impact of different aspects on system performance, together with still open issues. [ABSTRACT FROM AUTHOR]

• A review about the use and evolution of IoT solutions in smart health care is presented. • The work discusses the role of AI in the internet of medical things context. • Particular attention is given to solutions and approaches exploiting fog and edge computing for running AI algorithms. In recent times, we assist to an ever growing diffusion of smart medical sensors and Internet of things devices that are heavily changing the way healthcare is approached worldwide. In this context, a combination of Cloud and IoT architectures is often exploited to make smart healthcare systems capable of supporting near realtime applications when processing and performing Artificial Intelligence on the huge amount of data produced by wearable sensor networks. Anyway, the response time and the availability of cloud based systems, together with security and privacy, still represent critical issues that prevents Internet of Medical Things (IoMT) devices and architectures from being a reliable and effective solution to the aim. Lately, there is a growing interest towards architectures and approaches that exploit Edge and Fog computing as an answer to compensate the weaknesses of the cloud. In this paper, we propose a short review about the general use of IoT solutions in health care, starting from early health monitoring solutions from wearable sensors up to a discussion about the latest trends in fog/edge computing for smart health. [ABSTRACT FROM AUTHOR]

• Five state of the art subgraph isomorphism algorithms have been compared. • Eight different datasets publicly available have been used in our experimentation. • The datasets have been extracted from both synthetic and real data. • Eleven different typologies of synthetic graphs have been included in the synthetic part of the dataset. • The experimentation required an impressive time of 556 days. Graph matching algorithms are gaining more and more interest in the last years from different scientific communities; indeed, they allow comparing any kind of objects represented using their intrinsic structure, represented in terms of attributed relational graphs. The challenge is to make these algorithms able to provide solutions over huge graphs, with many thousands of nodes, and in a time that is adequate for practical applications; in this paper, we propose a comparison among the best performing algorithms available in the literature on a variety of very large graph databases used for performance assessment. The chosen datasets vary in terms of graph structure, size, density, presence of symmetric or repetitive substructures; this variability makes such datasets very challenging. The aim of this paper is to characterize the performance of the compared algorithms with respect to the typology, the size and other structural properties of the graphs; in this way, the user may consciously select the best suited algorithm for a given purpose. The results of an impressive experimentation that required 556 days of machine time are here presented and extensively discussed. [ABSTRACT FROM AUTHOR]

In Video Surveillance age, the monitoring activity, especially from unmanned vehicles, needs some degree of autonomy in the scenario interpretation. Video Analysis tasks are crucial for the target tracking and recognition; anyway, it would be desirable if a further level of understanding could provide a comprehensive, high-level scene description, by reflecting that human cognitive capability of providing a concise scene description that comes from the analysis of involved objects relationships and actions. This paper presents a smart system to identify mobile scene objects, such as people, vehicles, automatically, by analyzing the videos acquired by drones in flight, along with the activities they carried out, so as to depict what it happens in the scene from a high-level perspective. The system uses Artificial Vision methods to detect and track the mobile objects and the area where they move, and Semantic Web technologies to provide a high-level description of the scenario. Spatio/temporal relations among the tracked objects as well as simple object activities (events) are described. By semantic reasoning, the system is able to connect the simple activities into more complex activities, that better reflect a human-like description of a scenario portion. Tests conducted on several videos, showing scenarios set in different environments, return convincing results which affirm the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

• VF3-Light is a simplified version of the VF3 subgraph isomorphism algorithm. • The use of simpler heuristics makes VF3-Light more effective on some kinds of graphs. • An experimental evaluation shows good performance on several kinds of graphs. • VF3-Light performs well even with respect to other state-of-the-art algorithms. In this paper we introduce VF3-Light, a simplification of VF3, a recently introduced, general-purpose subgraph isomorphism algorithm. While VF3 has demonstrated to be very effective on several datasets, especially on very large and very dense graphs, we will show that on some classes of graphs, the full power of VF3 may become an overkill; indeed, by removing some of the heuristics used in it, and as a consequence also some of the data structures that are required by them, we obtain an algorithm (VF3-Light) that is actually faster. In order to provide a characterization of this modified algorithm, we have performed an evaluation using several publicly available graph datasets. Besides comparing VF3-Light with VF3, we have also included in the comparison other recent algorithms that are rated among the fastest in the state of the art. [ABSTRACT FROM AUTHOR]

Visual tracking supported by unmanned aerial vehicles (UAVs) has generated a lot of interest in recent years, especially in application domains such as surveillance, search for missing persons and traffic monitoring. The major challenges in visual tracking with small UAVs arise in the form of target representation, target appearance change, target detection and localization in real time computation. Reliable target detection depends on factors such as occlusions, image noise, illumination and pose changes, or image blur that may compromise the object labeling. To mitigate these issues, this paper proposes a hybrid solution: along with the tracked objects, scenes are completely depicted by adding contextual information, i.e., data describing places, natural features, or in general points of interest. Each scenario indeed is semantically described by ontological statements that define the context and then, by inference, support the object tracking task in the object identification and labeling. The synergy between the tracking methods and semantic modeling can bridge the object labeling gap, enhancing the scene understanding and awareness when alarming situations are discovered. Experimental results are promising and confirm the applicability of the proposed framework in supporting drones in object identification and critical situation detection tasks. [ABSTRACT FROM AUTHOR]

Graph matching is essential in several fields that use structured information, such as biology, chemistry, social networks, knowledge management, document analysis and others. Except for special classes of graphs, graph matching has in the worst-case an exponential complexity; however, there are algorithms that show an acceptable execution time, as long as the graphs are not too large and not too dense. In this paper we introduce a novel subgraph isomorphism algorithm, VF3, particularly efficient in the challenging case of graphs with thousands of nodes and a high edge density. Its performance, both in terms of time and memory, has been assessed on a large dataset of 12,700 random graphs with a size up to 10,000 nodes, made publicly available. VF3 has been compared with four other state-of-the-art algorithms, and the huge experimentation required more than two years of processing time. The results confirm that VF3 definitely outperforms the other algorithms when the graphs become huge and dense, but also has a very good performance on smaller or sparser graphs. [ABSTRACT FROM PUBLISHER]

The Graph Edit Distance (GED) is a flexible measure of dissimilarity between graphs which arises in error-correcting graph matching. It is defined from an optimal sequence of edit operations (edit path) transforming one graph into another. Unfortunately, the exact computation of this measure is NP-hard. In the last decade, several approaches were proposed to approximate the GED in polynomial time, mainly by solving linear programming problems. Among them, the bipartite GED received much attention. It is deduced from a linear sum assignment of the nodes of the two graphs, which can be efficiently computed by Hungarian-type algorithms. However, edit operations on nodes and edges are not handled simultaneously, which limits the accuracy of the approximation. To overcome this limitation, we propose to extend the linear assignment model to a quadratic one. This is achieved through the definition of a family of edit paths induced by assignments between nodes. We formally show that the GED, restricted to the paths in this family, is equivalent to a quadratic assignment problem. Since this problem is NP-hard, we propose to compute an approximate solution by adapting two algorithms: Integer Projected Fixed Point method and Graduated Non Convexity and Concavity Procedure. Experiments show that the proposed approach is generally able to reach a more accurate approximation of the exact GED than the bipartite GED, with a computational cost that is still affordable for graphs of non trivial sizes. [ABSTRACT FROM AUTHOR]

More and more real-world applications, especially in cognitive robotics, require the running of different computer vision algorithms in parallel on board of embedded devices with limited GPU and memory resources. Multi-task learning, namely the usage of the same model to perform multiple classification and/or regression tasks by learning a shared low level representation, revealed to be a valid solution to reduce computation and required memory space while preserving the accuracy. In this paper, we propose a solution for real-time user profiling based on a multi-task convolutional neural network (CNN) for gender, age, ethnicity and emotion recognition from face images. To find the best trade-off between accuracy and processing time, we evaluate three different architectures, specifically designed for the purpose, and backbones, based on MobileNet, ResNet and SENet, which include convolutional layers, residual blocks and attention modules that already demonstrated great potential in face analysis. We trained the multi-task neural network with a custom learning procedure, which solves the problems of missing labels, dataset imbalance and loss function imbalance through label masking, batch balancing and a custom weighted loss function; there are no other multi-task neural networks for face analysis that address all these challenges simultaneously.. The proposed solution demonstrated its effectiveness in the comparison with the corresponding single-task CNNs in terms of accuracy, processing time and memory space; in fact, the multi-task CNNs achieved a processing speed-up between 2.5 and 4 times and a reduction of the memory space between 2 and 4 times, while preserving the accuracy. Moreover, the useful insights that arise from the experiments allow to choose a solution for face analysis easily integrable into real applications on smart cameras and embedded systems and most suited for the specific application constraints in terms of computational resources. [ABSTRACT FROM AUTHOR]

The Indirect Immunofluorescence (IIF) protocol applied on Human Epithelial type 2 (HEp-2) cells is the current gold standard for the Antinuclear Antibody (ANA) test. The formulation of the diagnosis requires the visual analysis of a patient’s specimen under a fluorescence microscope in order to recognize the cells’ staining pattern which could be related to a connective tissue disease. This analysis is time consuming and error prone, thus in the recent past we have witnessed a growing interest in the pattern recognition scientific community directed at the development of methods for supporting this complex task. The main driver of the interest towards this problem is represented by the series of international benchmarking initiatives organized in the last four years that allowed dozens of research groups to propose innovative methodologies for HEp-2 cells’ staining pattern classification. In this paper we update the state of the art on HEp-2 cells and specimens classification, by analyzing the performance achieved by the methods participating the contest on Performance Evaluation of IIF Image Analysis Systems, hosted by the 22nd edition of the International Conference on Pattern Recognition ICPR 2014, and to the Executable Thematic Special Issue of Pattern Recognition Letters on Pattern Recognition Techniques for IIF Images Analysis, and by highlighting the trends in the design of the best performing methods. [ABSTRACT FROM AUTHOR]

The Anti-Nuclear Antibodies (ANA) test using Human Epithelial (HEp-2) cells has been the gold standard to identify the presence of Connective Tissue Diseases (CTD) such us Systemic Lupus Erythomatus (SLE). As the ANA test is time consuming, labor intensive and subjective, there has been an on-going effort to develop image-based Computer Aided Diagnosis (CAD) systems. This paper discusses the current progress and challenges in this field and will highlight areas which require more attention. [ABSTRACT FROM AUTHOR]

In this paper we present a vision based method for counting the number of persons which cross a virtual line. The method analyzes the video stream acquired by a camera mounted in a zenithal position with respect to the counting line, allowing to determine the number of persons that cross the virtual line and providing the crossing direction for each person. The proposed approach has been specifically designed to achieve high accuracy and computational efficiency, so as to allow its adoption in real scenarios. An extensive evaluation of the method has been carried out taking into account the main factors that may impact on the counting performance and, in particular, the acquisition technology (traditional RGB camera and depth sensor), the installation scenario (indoor and outdoor), the density of the people flow (isolated people and groups of persons), the acquisition frame rate, and the image resolution. We have also analyzed the combination of the outputs obtained from the RGB and depth sensors as a way to improve the counting performance. The experimental results confirm the effectiveness of the proposed method, especially when combining RGB and depth information, and the tests over three different CPU architectures demonstrate the possibility of deploying the method both on high-end servers for processing in parallel a large number of video streams and on low power CPUs as those embedded on commercial smart cameras. [ABSTRACT FROM AUTHOR]

Objective: This paper presents benchmarking results of human epithelial type 2 (HEp-2) interphase cell image classification methods on a very large dataset. The indirect immunofluorescence method applied on HEp-2 cells has been the gold standard to identify connective tissue diseases such as systemic lupus erythematosus and Sjögren's syndrome. However, the method suffers from numerous issues such as being subjective, time consuming and labor intensive. This has been the main motivation for the development of various computer-aided diagnosis systems whose main task is to automatically classify a given cell image into one of the predefined classes.Methods and Material: The benchmarking was performed in the form of an international competition held in conjunction with the International Conference of Image Processing in 2013: fourteen teams, composed of practitioners and researchers in this area, took part in the initiative. The system developed by each team was trained and tested on a very large HEp-2 cell dataset comprising over 68,000 images of HEp-2 cell. The dataset contains cells with six different staining patterns and two levels of fluorescence intensity. For each method we provide a brief description highlighting the design choices and an in-depth analysis on the benchmarking results.Results: The staining pattern recognition accuracy attained by the methods varies between 47.91% and slightly above 83.65%. However, the difference between the top performing method and the seventh ranked method is only 5%. In the paper, we also study the performance achieved by fusing the best methods, finding that a recognition rate of 85.60% is reached when the top seven methods are employed.Conclusions: We found that highest performance is obtained when using a strong classifier (typically a kernelised support vector machine) in conjunction with features extracted from local statistics. Furthermore, the misclassification profiles of the different methods highlight that some staining patterns are intrinsically more difficult to recognize. We also noted that performance is strongly affected by the fluorescence intensity level. Thus, low accuracy is to be expected when analyzing low contrasted images. [ABSTRACT FROM AUTHOR]

In this paper we propose a novel method for the detection of audio events for surveillance applications. The method is based on the bag of words approach, adapted to deal with the specific issues of audio surveillance: the need to recognize both short and long sounds, the presence of a significant noise level and of superimposed background sounds of intensity comparable to the audio events to be detected. In order to test the proposed method in complex, realistic scenarios, we have built a large, publicly available dataset of audio events. The dataset has allowed us to evaluate the robustness of our method with respect to varying levels of the Signal-to-Noise Ratio; the experimentation has confirmed its applicability under real world conditions, and has shown a significant performance improvement with respect to other methods from the literature. [ABSTRACT FROM AUTHOR]

Highlights: [•] We distinguish between mitotic and interphase HEp-2 cells. [•] We distinguish positive and negative mitotic cells. [•] We apply a cascade recognition approach using a set of features tailored for the task. [•] We adopt a training methodology specifically devised for skewed samples distributions. [•] The approach has been evaluated on a public reference dataset of HEp-2 cells. [Copyright &y& Elsevier]

Abstract: Indirect Immunofluorescence (IIF) images are increasingly being used for the diagnosis of autoimmune diseases. However, the analysis of this kind of images has until now reached a comparatively low level of automation, if compared with other medical imaging techniques. The Special Issue on the Analysis and Recognition of Indirect Immunofluorescence Images of the Pattern Recognition journal aims at providing a comprehensive evaluation of the state of the art for the staining pattern classification problem, through the adoption of a common experimental protocol and the testing of all the methods on a publicly available dataset. This paper will present both a survey of the papers accepted for the special issue, highlighting their original aspects, and a detailed comparison and discussion of the corresponding experimental results, in order to assess which are the advantages and disadvantages of each approach. [Copyright &y& Elsevier]

In this paper, we report on the first edition of the HEp-2 Cells Classification contest, held at the 2012 edition of the International Conference on Pattern Recognition, and focused on indirect immunofluorescence (IIF) image analysis. The IIF methodology is used to detect autoimmune diseases by searching for antibodies in the patient serum but, unfortunately, it is still a subjective method that depends too heavily on the experience and expertise of the physician. This has been the motivation behind the recent initial developments of computer aided diagnosis systems in this field. The contest aimed to bring together researchers interested in the performance evaluation of algorithms for IIF image analysis: 28 different recognition systems able to automatically recognize the staining pattern of cells within IIF images were tested on the same undisclosed dataset. In particular, the dataset takes into account the six staining patterns that occur most frequently in the daily diagnostic practice: centromere, nucleolar, homogeneous, fine speckled, coarse speckled, and cytoplasmic. In the paper, we briefly describe all the submitted methods, analyze the obtained results, and discuss the design choices conditioning the performance of each method. [ABSTRACT FROM AUTHOR]

The paper presents a method for estimating the number of moving people in a scene for video surveillance applications. The method performance has been characterized on the public database used for the PETS 2009 and 2010 international competitions; the proposed method has been compared, on the same database, with the PETS competitions participants. The system exhibits a high accuracy, and revealed to be so fast that it can be used in real time surveillance applications. The rationale of the method lies on the extraction of suited scale-invariant feature points and the successive selection among them of the moving ones, under the hypothesis that the latter are associated to moving people. The perspective distortions are taken into account by dividing the input frames into smaller horizontal zones, each having (approximately) the same perspective effects. Therefore, the evaluation of the number of people is separately carried out for each zone, and the results are summed up. The most important peculiarity of the proposed method is the availability of a simple training procedure using a brief video sequence that shows a person walking around in the scene; the procedure automatically evaluates all the parameters needed by the system, thus making the method particularly suited for end-user applications. [ABSTRACT FROM AUTHOR]

This paper presents a novel algorithm for the removal of reflections generated by objects on reflecting floors. The algorithm uses both chromatic properties of the reflections and geometrical constraints on their positions; however, it does not make use of a model of the reflected objects, and so it can be applied to scenes containing several kinds of objects (e.g., people, baggage, animals, vehicles, etc.). The proposed method has been validated by an extensive set of experiments on a large video database. In these experiments, the method has been compared to two other recent reflection removal algorithms. The experimental results show that the proposed method is fast and effective, both in absolute terms and in comparison with the other algorithms. [ABSTRACT FROM AUTHOR]

In some Computer Vision applications there is the need for grouping, in one or more clusters, only a part of the whole dataset. This happens, for example, when samples of interest for the application at hand are present together with several noisy samples. In this paper we present a graph-based algorithm for cluster detection that is particularly suited for detecting clusters of any size and shape, without the need of specifying either the actual number of clusters or the other parameters. The algorithm has been tested on data coming from two different computer vision applications. A comparison with other four state-of-the-art graph-based algorithms was also provided, demonstrating the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

A common problem encountered in structural pattern recognition is the difficulty of constructing classification models or rules from a set of examples, due to the complexity of the structures needed to represent the patterns. In this paper, we present an extension of a method for structural learning. The goal of the method is to find descriptions which are general (in other words, are successfully applicable to recognize objects different from the ones in the training set), preserving at the same time their discrimination ability. This method has been applied to predictive toxicology evaluation, that is the inference of the cancerogenic characteristics of chemical compounds. [ABSTRACT FROM AUTHOR]

Abstract: One of the main difficult problem in video analysis is to track moving objects during a video sequence, especially in presence of occlusions. Unfortunately, almost all the different approaches work on a pixel-by-pixel basis, yielding them unusable in real-time situations as well as not much expressive at a semantic level of the video. We present in this paper a novel method of tracking objects through occlusions that exploits the wealth of information due to the spatial coherence between pixels, using a graph-based, multi-resolution representation of the moving regions. The experimental results show that the approach is promising. [Copyright &y& Elsevier]

Shot Change Detection (SCD) in MPEG coded videos is a complex and still open research problem whose interest is growing up more and more due to the diffusion of Video Databases and Digital Libraries. Techniques providing fully satisfactory performances on complex video domains are not yet available even if a number of proposals exist; such proposals show very often to be complementary in their results. In this context, the Authors investigated the use of Multi-Expert Systems (MES) for approaching the SCD problem. In the present paper, we propose and discuss a strategy to select the SCD techniques to be combined and a method for choosing an effective combining rule. In order to assess the performance of the proposed MES, we set up a database that is significantly wider than the ones commonly used in the field. Experimental results demonstrate that the proposed system performs better than each of the single SCD technique considered. [ABSTRACT FROM AUTHOR]

Mammography is a not invasive diagnostic technique widely used for early cancer detection in women breast. A significant visual clue of the disease is the presence of clusters of microcalcifications. The automatic recognition of malignant clusters of microcalcifications, which could be very helpful for diagnostic purposes, is a very difficult task because of the small size of the microcalcifications and of the poor quality of the mammographic images.In this paper we propose a novel approach for classifying clusters of microcalcifications, based on a Multiple Expert System; such system aggregates several experts, some of which are devoted to classify the single microcalcifications while others are aimed to classify the cluster considered as a whole. The final output results from the suitable combination of the two groups of experts. The tests performed on a standard database of 40 mammographic images have confirmed the effectiveness of the approach. [Copyright &y& Elsevier]