Skin cancers are the most cancers diagnosed worldwide, with an estimated > 1.5 million new cases in 2020. Use of computer-aided diagnosis (CAD) systems for early detection and classification of skin lesions helps reduce skin cancer mortality rates. Inspired by the success of the transformer network in natural language processing (NLP) and the deep convolutional neural network (DCNN) in computer vision, we propose an end-to-end CNN transformer hybrid model with a focal loss (FL) function to classify skin lesion images. First, the CNN extracts low-level, local feature maps from the dermoscopic images. In the second stage, the vision transformer (ViT) globally models these features, then extracts abstract and high-level semantic information, and finally sends this to the multi-layer perceptron (MLP) head for classification. Based on an evaluation of three different loss functions, the FL-based algorithm is aimed to improve the extreme class imbalance that exists in the International Skin Imaging Collaboration (ISIC) 2018 dataset. The experimental analysis demonstrates that impressive results of skin lesion classification are achieved by employing the hybrid model and FL strategy, which shows significantly high performance and outperforms the existing work.

Precision farming technologies refer to a set of cutting-edge tools and strategies implemented to optimize the management of the plantation. Smart meter devices, IoT technologies, and Wireless Sensor Networks are only a few examples of the innovative systems increasingly employed from an Agriculture 4.0 point of view. Recent literature has paid close attention to the role of Artificial Intelligence (AI) and Deep Learning (DL) algorithms in helping farmers and improving soil productivity. In this regard, this paper presents the design of a Wireless Sensor Network based on low-cost, low-power PV-supplied sensor nodes able to acquire data regarding environmental conditions and soil parameters. Among all the implemented sensors, the most critical is the soil moisture sensors because of many issues related to cost, installation, reliability, and calibration. Thus, this paper proposes a deep learning approach based on Long Short-Term Memory (LSTM) networks to provide a virtual soil moisture sensor using only the data acquired by the other transducer installed on the node. Performances estimation of the virtual sensors and an in-depth comparison with other learning-based approaches have been presented in this paper to validate the effectiveness of the proposed soft sensing approach.

The recent development of highly automated machinery and intelligent industrial plants has increasingly enabled the continuous monitoring of their efficiency and condition, with the aim of maintaining high production efficiency and minimal malfunctions. Typical condition monitoring and fault detection applications are often achieved using acoustic and vibrational techniques, but the availability of distributed electrical measurements opens new opportunities for industrial fault detection with minimal impact on electrical systems. Even if artificial intelligence-based approaches can be used to model industrial equipment by means of measures made on electrical systems to which they are connected, machine learning algorithms have been demonstrated to be particularly adequate for this purpose due to the huge amount of data produced by interconnected sensors and devices. In this context, the aim of this work is to propose a new unsupervised analysis methodology for detecting anomalies in industrial machinery using electrical current values and other parameters measured on the power grid. The proposed framework is aimed at incorporating the advantages of Machine Learning algorithms and those of traditional analysis, optimizing their operation to improve performance and execution time; this also incorporates a methodology for analyzing the temporal dynamics of the anomaly based on Short-time Fourier transform to strengthen the performance of the detection. The results obtained showed excellent performance, both compared to the evaluations of a technical expert and to other methodologies used in the literature, with zero false positives detected in all datasets tested and a negligible number of undetected outlier events, less than 4% of the total in the datasets.

Skin cancer is one of the most threatening cancers, which spreads to the other parts of the body if not caught and treated early. During the last few years, the integration of deep learning into skin cancer has been a milestone in health care, and dermoscopic images are right at the center of this revolution. This review study focuses on the state-of-the-art automatic diagnosis of skin cancer from dermoscopic images based on deep learning. This work thoroughly explores the existing deep learning and its application in diagnosing dermoscopic images. This study aims to present and summarize the latest methodology in melanoma classification and the techniques to improve this. We discuss advancements in deep learning-based solutions to diagnose skin cancer, along with some challenges and future opportunities to strengthen these automatic systems to support dermatologists and enhance their ability to diagnose skin cancer.

The use of Micro-Electro-Mechanical Systems (MEMS)-based Inertial Measurement Units (IMUs) is widespread in many applications concerning monitoring, diagnostic, and/or controlling in navigation and transportation systems, as well as in low-cost applications for automotive and aeronautical fields. The data provided by the set of sensors typically present in IMUs, as accelerometers, gyroscopes, and magnetometers, are often used also for feeding suitable filtering and positioning algorithms able to correct the attitude and path of the vehicle on which they are installed or to provide the analytical redundancy needed for online diagnosis. Nevertheless, from one hand, the performance of low-cost MEMS-based IMUs are certified only under a small set of nominal operating conditions, and on the other hand, the filtering algorithms are often designed and verified under canonical additive noises. In this framework, the paper proposes a test-plan and a test setup for analyzing and characterizing the performance of filtering algorithms for positioning based on data coming from low-cost IMUs, and able to verify systematically the operating of such algorithms under real scenarios. Two kinds of very popular filtering algorithms have been considered, namely the Complementary filter and the Attitude and Heading Reference Systems (AHRS) Kalman filter, which belongs to two opposite approaches. The experimental results prove how the typical vibrations present in real scenarios can significantly affect the performance of such algorithms.

The environmental acoustic noise is considered as a big risk for today's population health. Consequently, the regulations in many countries commit themselves to control the exposition of people, imposing limits to the noise level. In the comparison between the measured value and the threshold, the uncertainty of the measured value has to be taken into account. In this paper, a procedure for the evaluation of the uncertainty of traffic noise measurements due to the variability of the measurand is proposed. A deep analysis of five bootstrap (normal, basic percentile, t-student, bias corrected percentile, and bias corrected and accelerated percentile) methods is performed to obtain accurate confidence intervals for the indicator Leq,A without necessity to make normal theory assumptions. From the comparison with the classical method (according to Guide to the Expression of Uncertainty in Measurement (ISO GUM)), the novel approach reveals to be more effective for estimating both the expected value and the uncertainty of the short-term equivalent sound pressure level when a large data set is not available.