Abstract
Morphing attacks are one of the many threats that are constantly affecting deep face recognition systems. It consists of selecting two faces from different individuals and fusing them into a final image that contains the identity information of both. In this work, we propose a novel regularisation term that takes into account the existent identity information in both and promotes the creation of two orthogonal latent vectors. We evaluate our proposed method (OrthoMAD) in five different types of morphing in the FRLL dataset and evaluate the performance of our model when trained on five distinct datasets. With a small ResNet-18 as the backbone, we achieve state-of-the-art results in the majority of the experiments, and competitive results in the others.

Abstract
The forecast of wind speed is prerequisite for wind power prediction, which is one of the most effective means of promoting wind power absorption. However, when modeling for wind speed sequences with different fluctuations, most existing researches ignore the influence of time scale of wind speed fluctuation period, let alone the low compatibility between training and testing samples that severely limit the training performance of forecasting model. To improve the accuracy of wind speed and wind power forecasting, an ultra-short-term wind speed forecasting model based on time scale recognition and dynamic adaptive modeling is proposed in this paper. First, a series of wind processes are divided from the historical wind speed sequence according to the natural variation characteristics of wind speed. Second, we divide all the wind processes into two patterns based on their time scale, and an SVC model with input features extracted from meteorological data is built to identify the time scale of the current wind process. Third, for a specifically identified wind process, the complex network algorithm is applied in data screening to select high compatible training samples to train the forecast model dynamically for current input. Simulation indicates that the proposed approach presents higher accuracy than benchmark models using the same forecasting algorithms but without considering the time scale and data screening.

Abstract
Chest radiography is one of the most common medical imaging modalites. However, chest radiography interpretation is a complex task that requires significant expertise. As such, the development of automatic systems for pathology detection has been proposed in literature, particularly using deep learning. However, these techniques suffer from a lack of explainability, which hinders their adoption in clinical scenarios. One technique commonly used by radiologists to support and explain decisions is to search for cases with similar findings for direct comparison. However, this process is extremely time-consuming and can be prone to confirmation bias. Automatic image retrieval methods have been proposed in literature but typically extract features from the whole image, failing to focus on the lesion in which the radiologist is interested. In order to overcome these issues, a novel framework LXIR for lesion-based image retrieval is proposed in this study, based on a state of the art object detection framework (YOLOv5) for the detection of relevant lesions as well as feature representation of those lesions. It is shown that the proposed method can successfully identify lesions and extract features which accurately describe high-order characteristics of each lesion, allowing to retrieve lesions of the same pathological class. Furthermore, it is show that in comparison to SSIM-based retrieval, a classical perceptual metric, and random retrieval of lesions, the proposed method retrieves the most relevant lesions 81% of times, according to the evaluation of two independent radiologists, in comparison to 42% of times by SSIM.

Abstract
The atmospheric electric field is a key characteristic of the Earth system. Despite its relevance, oceanic measurements of the atmospheric electric field are scarce, as typically oceanic measurements tend to be focused on ocean properties rather than on the atmosphere above. This motivated the set-up of an innovative campaign on board the sail ship NRP Sagres focused on the measurement of the atmospheric electric field in the marine boundary layer. This paper describes the monitoring system that was developed to measure the atmospheric electric field during the planned circumnavigation expedition of the sail ship NRP Sagres.

Abstract
n-Alkanes and long-chain alcohols (LCOH) have been used as faecal markers to assess the feeding behaviour of both wild and domestic herbivore species. However, their chemical analysis is time-consuming and expensive, making it necessary to develop more expeditious methodologies to evaluate concentrations of these markers. This work aimed to evaluate the use of Fourier Transform Infrared Spectroscopy (FTIR) technology in the near infrared (NIR) and mid infrared (MIR) intervals, for the determination of n-alkane and LCOH concentrations of different plant species and faecal samples of domestic herbivores. Spectra of 33 feed samples, namely L. perenne, T. repens, U. gallii, short heathers (mixture of Erica spp. and Calluna vulgaris), improved pasture grasses (mixture of L. perenne and A. capillaris), heath grasses (mixture of P. longifolium and A. curtissii), improved pasture species (mixture of L. perenne, T. repens and A. capillaris) and herbaceous species (mixture of all herbaceous species found in the plot)) and 181 faecal samples (cattle and horses) were recorded. In order to develop calibrations for the prediction of n-alkanes and LCOH concentrations, partial least squares (PLS) regression was used. Regarding the models developed for plant species, the best results were observed for the calibrations using NIR. The best external validation coefficients of determination (R(2)v) obtained were 0.90 and 0.79 for LCOH and n-alkanes, respectively. For faecal samples, in the NIR interval, results indicate similar external validation predictions (R(2)v) for both animal species (0.64). On the contrary, in the MIR interval, differences between cattle (0.70) and horses (0.57) faecal samples in R(2)v were observed. Regarding the models created for both animal species faeces, LCOH (C-26-OH and C-30-OH concentrations ranging from 713.3 to 4451.9 mg/kg DM, respectively; R(2)v values ranging from 0.72 to 0.95) and n-alkanes (C31 and C33 concentrations ranging from 112.8 to 643.2 mg/kg DM, respectively; R(2)v values ranging from 0.19 to 0.90) present in higher concentrations tended to be those with better estimates. Results obtained suggest that the selection of the technique to be used may depend on the type of matrix, being the homogeneity of the matrices one of the most important factors for its success. In order to improve the accuracy and robustness of the models created for the estimation of the concentrations of these markers using these methodologies, the database (greater variability) used for the calibrations of these models must be increased.

Abstract
This chapter aims to contribute to a better understanding of the role of websites in business internationalization by exploring how website overall objectives and their coherence with website strategies support website internationalization effectiveness. It provides empirical evidence on the experiences of Portuguese companies shared by 20 managers of large companies and SMEs of various activity sectors. Results show the importance of a clear website strategy (e.g., clear objectives and coherent tactics) for an effective role in internationalization. Findings also confirm that, while many managers are skeptic about the effectiveness of websites as an internationalization touchpoint, namely due to sector characteristics (e.g., type of customers, type of products/services), the website is perceived as an essential tool for reaching, attracting, and involving international customers, supporting other communication instruments such as participating in international fairs and sales force.