Visual Odometry (VO) and Visual SLAM (VSLAM) systems often struggle in low-light and dark environments due to the lack of robust visual features. In this paper, we propose a novel active illumination framework to enhance the performance of VO and V-SLAM algorithms in these challenging conditions. The developed approach dynamically controls a moving light source to illuminate highly textured areas, thereby improving feature extraction and tracking. Specifically, a detector block, which incorporates a deep learning-based enhancing network, identifies regions with relevant features. Then, a pan-tilt controller is responsible for guiding the light beam toward these areas, so that to provide information-rich images to the ego-motion estimation algorithm. Experimental results on a real robotic platform demonstrate the effectiveness of the proposed method, showing a reduction in the pose estimation error up to 75 % with respect to a traditional fixed lighting technique.

Visual Inertial Odometry (VIO) and Simultaneous Localization and Mapping (SLAM) have experienced increasing interest in both the consumer and racing automotive sectors in recent decades. With the introduction of novel neuromorphic vision sensors, it is now possible to accurately localize a vehicle even under complex environmental conditions, leading to an improved and safer driving experience. In this paper, we propose MA-VIED, a large-scale driving dataset that collects race track-like loops, maneuvers, and standard driving scenarios, all bundled in a rich sensory dataset. MA-VIED provides highly accurate IMU data, standard and event camera streams, and RTK position data from a dual GPS antenna, both of which are hardware-synchronized with all cameras and IMU data. In addition, we collect accurate wheel odometry data and other data from the vehicle’s CAN bus. The dataset contains 13 sequences collected in urban, suburban, and racetrack-like environments with varying lighting conditions and driving dynamics. We provide ground-truth RTK data for algorithms evaluation and the calibration sequences for both IMU and cameras. We then present three tests to demonstrate how MA-VIED can be suitable for monocular VIO applications, using state-of-the-art VIO algorithms and an EKF-based sensor fusion solution. The experimental results show that MA-VIED can support the development and prototyping of novel automotive-oriented frame and event-based monocular VIO algorithms.

Human-robot interaction and affective computing are cross-disciplinary fields whose connections are being further explored with the proliferation of humanoid robots that assist people in their daily activities. The proposed study aims to assess emotion recognition accuracy based on facial image acquisition, recorded by the camera sensor of the humanoid robot Pepper, during a human-robot interaction scenario. The emotions of the individuals involved were elicited by viewing five different video clips, each corresponding to different target emotions (e.g., happiness, sadness, anger, surprise, and neutrality). Emotion evaluation was carried out using the Pepper integrated classification module and convolutional neural networks, and the obtained results were compared to determine the best emotion classification accuracy. The test group was diverse, particularly in terms of participant age, including a significant number of elderly individuals who were not familiar with robot interaction. This diversity also allowed for an examination of user acceptability in the context of the human-robot interaction experience.

Abstract The performances of aircraft braking control systems are strongly influenced by the tire friction force experienced during the braking phase. The availability of an accurate estimate of the current airstrip characteristics is a recognized issue for developing optimized braking control schemes. The study presented in this paper is focused on the robust online estimation of the airstrip characteristics from sensory data usually available on an aircraft. In order to capture the nonlinear dependency of the current best slip on sequential slip-friction measurements acquired during the braking maneuver, multilayer perceptron (MLP) approximators have been proposed. The MLP training is based on a synthetic data set derived from a widely used tire–road friction model. In order to achieve robust predictions, MLP architectures based on the drop-out mechanism have been applied not only in the offline training phase but also during the braking. This allowed to online compute a confidence interval measure for best friction estimate that has been exploited to refine the estimation via Kalman Filtering. Open loop and closed loop simulation studies in 15 representative airstrip scenarios (with multiple surface transitions) have been performed to evaluate the performance of the proposed robust estimation method in terms of estimation error, aircraft braking distance, and time, together with a quantitative comparison with a state-of-the-art benchmark approach.

Natural language video description (NLVD) has recently received strong interest in the computer vision, natural language processing (NLP), multimedia, and autonomous robotics communities. The state-of-the-art (SotA) approaches obtained remarkable results when tested on the benchmark datasets. However, those approaches poorly generalize to new datasets. In addition, none of the existing works focus on the processing of the input to the NLVD systems, which is both visual and textual. In this paper, an extensive study is presented to deal with the role of the visual input, evaluated with respect to the overall NLP performance. This is achieved by performing data augmentation of the visual component, applying common transformations to model camera distortions, noise, lighting, and camera positioning that are typical in real-world operative scenarios. A t-SNE-based analysis is proposed to evaluate the effects of the considered transformations on the overall visual data distribution. For this study, the English subset of the Microsoft Research Video Description (MSVD) dataset is considered, which is used commonly for NLVD. It was observed that this dataset contains a relevant amount of syntactic and semantic errors. These errors have been amended manually, and the new version of the dataset (called MSVD-v2) is used in the experimentation. The MSVD-v2 dataset is released to help to gain insight into the NLVD problem.

Among the main challenges in robotics, target-driven visual navigation has gained increasing interest in recent years. In this task, an agent has to navigate in an environment to reach a user specified target, only through vision. Recent fruitful approaches rely on deep reinforcement learning, which has proven to be an effective framework to learn navigation policies. However, current state-of-the-art methods require to retrain, or at least fine-tune, the model for every new environment and object. In real scenarios, this operation can be extremely challenging or even dangerous. For these reasons, we address generalization in target-driven visual navigation by proposing a novel architecture composed of two networks, both exclusively trained in simulation. The first one has the objective of exploring the environment, while the other one of locating the target. They are specifically designed to work together, while separately trained to help generalization. In this article, we test our agent in both simulated and real scenarios, and validate its generalization capabilities through extensive experiments with previously unseen goals and unknown mazes, even much larger than the ones used for training.

In cancer research, robustness of a complex biochemical network is one of the most relevant properties to investigate for the development of novel targeted therapies. In cancer systems biology, biological networks are typically modeled through Ordinary Differential Equation (ODE) models. Hence, robustness analysis consists in quantifying how much the temporal behavior of a specific node is influenced by the perturbation of model parameters. The Conditional Robustness Algorithm (CRA) is a valuable methodology to perform robustness analysis on a selected output variable, representative of the proliferation activity of cancer disease. Here we introduce our new freely downloadable software, the CRA Toolbox. The CRA Toolbox is an Object-Oriented MATLAB package which implements the features of CRA for ODE models. It offers the users the ability to import a mathematical model in Systems Biology Markup Language (SBML), to perturb the model parameter space and to choose the reference node for the robustness analysis. The CRA Toolbox allows the users to visualize and save all the generated results through a user-friendly Graphical User Interface (GUI). The CRA Toolbox has a modular and flexible architecture since it is designed according to some engineering design patterns. This tool has been successfully applied in three nonlinear ODE models: the Prostate-specific Pten−/− mouse model, the Pulse Generator Network and the EGFR-IGF1R pathway. The CRA Toolbox for MATLAB is an open-source tool implementing the CRA to perform conditional robustness analysis. With its unique set of functions, the CRA Toolbox is a remarkable software for the topological study of biological networks. The source and example code and the corresponding documentation are freely available at the web site: http://gitlab.ict4life.com/SysBiOThe/CRA-Matlab .

Obstacle Detection is a central problem for any robotic system, and critical for autonomous systems that travel at high speeds in unpredictable environment. This is often achieved through scene depth estimation, by various means. When fast motion is considered, the detection range must be longer enough to allow for safe avoidance and path planning. Current solutions often make assumption on the motion of the vehicle that limit their applicability, or work at very limited ranges due to intrinsic constraints. We propose a novel appearance-based Object Detection system that is able to detect obstacles at very long range and at a very high speed (~ 300Hz), without making assumptions on the type of motion. We achieve these results using a Deep Neural Network approach trained on real and synthetic images and trading some depth accuracy for fast, robust and consistent operation. We show how photo-realistic synthetic images are able to solve the problem of training set dimension and variety typical of machine learning approaches, and how our system is robust to massive blurring of test images.

A robust gesture recognition system is an essential component in many human-computer interaction applications. In particular, the widespread adoption of portable devices and the diffusion of autonomous systems with limited power and load capacity has increased the need of developing efficient recognition algorithms which operates on video streams recorded from low cost devices and which can cope with the challenging issue of point of view changes. A further challenge arises as different users tend to perform the same gesture with different styles and speeds. Thus a classifier trained with gestures data of certain set of users may work poorly when data from other users are being processed. However, as often a mobile device or a robot are intended to be used by a single or by a small group of people, it would be desirable to have a gesture recognition system designed specifically for these users. In this paper we introduce a novel approach to face the problems of view-invariance and user personalization in the context of gesture interaction systems. More specifically, we propose a domain adaptation framework based on a feature space augmentation approach operating on robust view-invariant Self Similarity Matrix descriptors. To prove the effectiveness of our method a dataset corresponding to 17 users performing 10 different gestures under 3 point of views is collected and an extensive experimental evaluation is performed.

Following recent works on HRI for UAVs, we present a gesture recognition system which operates on the video stream recorded from a passive monocular camera installed on a quadcopter. While many challenges must be addressed for building a real-time vision-based gestural interface, in this paper we specifically focus on the problem of user personalization. Different users tend to perform the same gesture with different styles and speed. Thus, a system trained on visual sequences depicting some users may work poorly when data from other people are available. On the other hand, collecting and annotating many user-specific data is time consuming. To avoid these issues, in this paper we propose a personalized gestural interface. We introduce a novel transfer learning algorithm which, exploiting both data downloaded from the web and gestures collected from other users, permits to learn a set of person-specific classifiers. We integrate the proposed gesture recognition module into a HRI system with a flying quadrotor robot. In our system first the UAV localizes a person and individuates her identity. Then, when a user performs a specific gesture, the system recognizes it adopting the associated user-specific classifier and the quadcopter executes the corresponding task. Our experimental evaluation demonstrates that the proposed personalized gesture recognition solution is advantageous with respect to generic ones.

As researchers are striving for developing robotic systems able to move into the `the wild', the interest towards novel learning paradigms for domain adaptation has increased. In the specific application of semantic place recognition from cameras, supervised learning algorithms are typically adopted. However, once learning has been performed, if the robot is moved to another location, the acquired knowledge may be not useful, as the novel scenario can be very different from the old one. The obvious solution would be to retrain the model updating the robot internal representation of the environment. Unfortunately this procedure involves a very time consuming data-labeling effort at the human side. To avoid these issues, in this paper we propose a novel transfer learning approach for place categorization from visual cues. With our method the robot is able to decide automatically if and how much its internal knowledge is useful in the novel scenario. Differently from previous approaches, we consider the situation where the old and the novel scenario may differ significantly (not only the visual room appearance changes but also different room categories are present). Importantly, our approach does not require labeling from a human operator. We also propose a strategy for improving the performance of the proposed method by fusing two complementary visual cues. Our extensive experimental evaluation demonstrates the advantages of our approach on several sequences from publicly available datasets.

The place recognition module is a fundamental component in SLAM systems, as incorrect loop closures may result in severe errors in trajectory estimation. In the case of appearance-based methods the bag-of-words approach is typically employed for recognizing locations. This paper introduces a novel algorithm for improving loop closures detection performance by adopting a set of visual words weights, learned offline accordingly to a discriminative criterion. The proposed weights learning approach, based on the large margin paradigm, can be used for generic similarity functions and relies on an efficient online leaning algorithm in the training phase. As the computed weights are usually very sparse, a gain in terms of computational cost at recognition time is also obtained. Our experiments, conducted on publicly available datasets, demonstrate that the discriminative weights lead to loop closures detection results that are more accurate than the traditional bag-of-words method and that our place recognition approach is competitive with state-of-the-art methods.

Background. With the publication of large studies from different health systems comparing survival probabilities, cancer registries are increasingly involved in clinical evaluation research. The changing role of registries strictly depends on the integration between the oncology system and proper information technology (IT) tools. IT is fundamental to improving validity and timeliness of data diffusion when both the number of sources linked and the number of variables registered are on the rise. Aims. In this paper, we present a modern web-based management system that allows to integrate different sources, validate and elaborate data thus providing a new evaluation system for the oncology network based on cancer registries. Materials and methods. We developed a Web 2.0 management system for the Umbria Cancer Registry (S.G.RTUP) based on AMPAX technology (Apache, Mysql, PHP, Ajax and XML) and object-oriented programming. ISO/IEC 27001:2005 standard is followed to ensure security access to the information. The S.G.RTUP architecture is modular and extensible and information consistency is guaranteed by entity-relationship principles. Cancer sites, topology, morphology, and behavior are coded according to the International Classification of Diseases. Classical epidemiological indices for a cancer registry are implemented: incidence, mortality, years of potential life lost, and cumulative risk. S.G.RTUP has tools to prepare data for trend analysis and relative survival analysis. Geographical analysis is also implemented. Results. S.G.RTUP is integrated with the Oncology Network and gives timely epidemiological indices for evaluation of oncological activities. The registration system that we developed can effectively manage different data sources. Automatic importing of routinely available data from pathology archives, screening services, and hospital discharge records will reduce the time needed to produce data and will also allow the expansion of registered information. Several services for data visualization and statistical analysis are implemented. A geographic information system based on Google maps API is used for geolocalization of cases and map plotting of incidence and mortality rates. We implemented Besag York and Mollie's algorithm for real-time smoothed maps. All services can be dynamically performed over a subset of data that the user can select through an innovative filtering system. Discussion and conclusion. IT contributed to shortening all phases of cancer registration, including linkage with external sources, coding, quality control, data management and analysis and publication of results. Integration in the oncology network and secure Web access allowed us to design with clinicians innovative population-based collaborative studies. Our geographic analysis system enables us to develop sophisticated dynamic geostatistic tools.

In this paper we study the state estimation problem for a basic model of gene expression. The model describes the interaction between mRNA and protein(s) within the transcription- translation process. We first study some basic dynamic properties of the model, namely the existence of a unique equilibrium point, its stability and local observability. Then, we address the problem of estimating mRNA concentration by only measuring total protein concentration, and we propose a solution to it by means of different state estimation schemes. Finally, by means ofinsilicoexperiments, we compare the performance of the proposed structures.

In this paper, aHinfincontrol technique is proposed for the management of a supply chain (SC) model linearized about nominal operating conditions. The performance considered is a weightedHinfinnorm comprising the inventory and the orders placed by each site, hence controlling the bullwhip effect together with local management costs. It is shown how the optimization of local costs at each site is related to the performance of the whole chain and a decentralized control methodology is proposed based on this relation.

In this paper we describe an efficient software architecture for object-tracking, based on a stereoscopic vision system, that has been applied to a mobile robot controlled by a PC. After analyzing the epipolar rectification required to correct the original stereo-images, it is described a new valid and efficient algorithm for ball recognition (indeed circle detection) which is able to work in different lighting conditions and in a manner faster than some modified versions of Circle Hough Transform. Then, we show that stereo vision, besides giving an optimum estimation of the 3D position of the object, is useful to remove lots of the false identifications of the ball, thanks to the advantages of epipolar constraint. Finally, we describe a new strategy for ball following, by a mobile robot, which is able to” look for” the object whenever it comes out of the cameras view, by taking advantage of a” block matching” method similar to that of MPEG Video.

In this paper we study optimal scheduling for a two part-type, symmetric manufacturing systems subject to non negligible setup times, and characterized by finite buffer capacity. The solution method relies on 1) restricting the possible control policies to those which respect some general necessary conditions of optimality, and on 2) the exploitation of such properties to introduce a two-dimensional sampled version of the original system. The symmetric assumptions allows to map the two-dimensional sampled model onto an equivalent scalar one. By means of such a scalar sampled model we give the analytical solution of the optimal control problem addressed.

Considers the use of Petri nets for the study of the deadlock problem for two classes of kanban manufacturing systems: kanban flow lines and reentrant kanban flow lines. In the Petri net framework deadlock can be easily characterized by means of the siphon concept. The first result of the paper is the formal proof that kanban flow lines are deadlock free, regardless of the initial marking of the underlying Petri net. Vice versa, for the class of reentrant kanban flow lines, where resource sharing occurs, by means of simple examples we show that deadlock can actually occur. Formal results for the latter class are given, characterizing some fundamental properties of "critical" siphons i.e., siphons that may become empty leading to deadlock.

We study the problem of dynamic part scheduling and kanban allocation on a pull manufacturing system, with finite inventory space and stochastic demand. It is assumed that a finite number of requests can be backlogged and requests finding the full system are dropped. The system is a single machine, two part-type, and the control objective is that of minimizing an infinite horizon discounted cost in the inventory and backlog level and demand losses. No setup times and costs are considered. The original scheduling and kanban allocation problem is reformulated as a "pure" scheduling problem for constant kanban systems. A continuous-time Markov decision problem is formulated, and solved by means of a uniformization procedure and discrete-time dynamic programming. Some initial analytical results are given, while the solution structure is derived numerically.

The problem of dynamic control of manufacturing systems is addressed, and a solution is proposed in terms of a state reconstruction algorithm, that allows to exactly compute system performance under perturbed values of some system parameters. The control scheme is able to react to disturbances acting on the system. It is applied to a multi part-type manufacturing system, with the purpose of dynamically allocating buffer capacities in order to maximize system throughput. Simulation results are presented to illustrate the effectiveness of the scheme.

The paper discusses the modelling issues for a two link flexible robot arm. The exact dynamic model of the robot arm is derived, and approximate forms suitable for the analysis and control design are proposed. The approach used here to derive an approximate finite dimensional model of the flexible robot arm is based on the Ritz expansion method. Simulations and experimental tests have been carried out to validate the approximate dynamic model of the robot.

A simple asymptotic observer is proposed for the estimation of the generalized velocities of a robotic manipulator. Singular perturbation theory is used in the stability analysis of the error dynamics. High gains are used in the cancellation of the nonlinearities characterizing the dynamic behavior of the robot. The accuracy of the proposed algorithm is illustrated in simulation runs and verified through some experimental tests.>