Search

Accurate measurement of soil moisture (SM) is crucial for understanding crop growing conditions, optimizing irrigation practices, and early detection of drought. Synthetic aperture radar (SAR) has proven effective in SM inversion for agricultural scenarios. Machine learning methods enable large-scale SM mapping by circumventing complex computations and exhibiting high nonlinear fitting capabilities. While previous studies have explored SM retrieval using SAR data and machine learning with dual-polarimetric (DP), compact-polarimetric (CP), or quad-polarimetric (FP) modes, a comprehensive comparative study of these polarization modes in crop scenario is lacking. In this study, we assessed SM inversion using three SAR polarimetric modes (DP, CP, and FP) in C-band across three crop types (wheat, corn, and soybean) using multi-year RADARSAT-2 data. Various polarimetric backscattering variables, polarimetric decomposition parameters, and vegetation indices under different polarimetric modes were extracted to construct the corresponding SAR feature sets. Four machine learning algorithms, including Bagged Decision Tree (BAGTREE), Random Forest Regression (RFR), Extreme Gradient Boosting (XGB), and Gaussian Process Regression (GPR), were used. Additionally, forward feature selection (FFS) procedure was employed to reduce redundant features and enhance accuracy. Results indicate that FP mode consistently demonstrated superior performance in SM retrieval, with CP mode slightly trailing behind, and DP mode yielding the least favourable outcomes. The FFS method consistently enhanced SM retrieval accuracy. Among the machine methods, RFR and GPR exhibited superior performance across all three crop scenarios and three polarimetric modes. Specifically, RFR achieved the best accuracy in the corn and soybean scenarios, with root mean square errors (RMSE) of 4.46 vol.% and 6.49 vol.%, respectively, while GPR excelled in the wheat scenario, with a RMSE of 4.29 vol.%. FFS outputs highlig

The current tourism landscape intertwines physical and digital realms, giving rise to smart tourism. This paradigm shift leverages technology, generating large amounts of data from sensors and mobile devices, offering vast opportunities for strategic tourism planning and management. Smart tourism destinations (STDs) are pivotal in this context, utilising technology to enhance visitor experiences and facilitate informed decision-making through proficient data management. Within this framework, open data emerges as a crucial aspect. This study aims to demonstrate how combining open data with a local business campaign data can innovate tourism management and foster smart ecosystems. This approach facilitates public-private collaboration and strengthens STD models at various administrative levels. A novel search engine was used to identify suitable open data for integration, employing word embeddings to perform semantic searches and enhance traditional data retrieval methods. The findings highlight the potential of open data in supporting the tourism industry. Integrating open and proprietary data not only enriches tourist experiences but also informs destination management strategies. The contributions of this work enable STDs to enhance the decision-making process by (i) demonstrating the feasibility of integrating retail data and open data for tourism management; (ii) describing an AI-based data search engine for discovering open data relevant to STDs; and (iii) presenting and discussing new findings on the use of retail data in conjunction with external open data in a real case study.

This study explores the enhancement of human-robot interaction (HRI) through multimodal emotional recognition within social robotics, using the humanoid robot Pepper as a testbed. Despite the advanced interactive capabilities of robots like Pepper, their ability to accurately interpret and respond to human emotions remains limited. This paper addresses these limitations by integrating visual, auditory, and textual analyses to improve emotion recognition accuracy and contextual understanding. By leveraging multimodal data, the study aims to facilitate more natural and effective interactions between humans and robots, particularly in assistive, educational, and healthcare settings. The methods employed include convolutional neural networks for visual emotion detection, audio processing techniques for auditory emotion analysis, and natural language processing for text-based sentiment analysis. The results demonstrate that the multimodal approach significantly enhances the robot’s interactive and empathetic capabilities. This paper discusses the specific improvements observed, the challenges encountered, and potential future directions for research in multimodal emotional recognition in HRI.

In the era of big SAR data, it is urgent to develop dynamic time series DInSAR processing procedures for near-real-time monitoring of landslides. However, the dense vegetation coverage in mountainous areas causes severe decorrelations, which demands high precision and efficiency of phase optimization processing. The common phase optimization using single-polarization SAR data cannot produce satisfactory results due to the limited statistical samples in some natural scenarios. The novel polarimetric phase optimization algorithms, however, have low computational efficiency, limiting their applications in large-scale scenarios and long data sequences. In addition, temporal changes in the scattering properties of ground features and the continuous increase of SAR data require dynamic phase optimization processing. To achieve efficient phase optimization for dynamic DInSAR time series analysis, we combine the Sequential Estimator (SE) with the Total Power (TP) polarization stacking method and solve it using eigen decomposition-based Maximum Likelihood Estimator (EMI), named SETP-EMI. The simulation and real data experiments demonstrate the significant improvements of the SETP-EMI method in precision and efficiency compared to the EMI and TP-EMI methods. The SETP-EMI exhibits an increase of more than 50% and 20% in highly coherent points for the real data compared to the EMI and TP-EMI, respectively. It, meanwhile, achieves approximately six and two times more efficient than the EMI and TP-EMI methods in the real data case. These results highlight the effectiveness of the SETP-EMI method in promptly capturing and analyzing evolving landslide deformations, providing valuable insights for real-time monitoring and decision-making.

A dataset of sub-daily C-band data, acquired with a ground-based synthetic aperture radar, has been used to study soil and vegetation dynamics during a complete growing season in a controlled agricultural test site. The data have been exploited to analyse the rate and sources of decorrelation in the scene, as well as the consequences of the observation conditions of a sub-daily satellite (with either low, medium or geosynchronous orbit): short revisit times, availability of multiple acquisitions during a single day, and shallow observations at some incidence angles. Repeat-pass coherence is found to be less affected by temporal decorrelation when the primary image is acquired during nighttime or the last hours predawn. Regarding the incidence angle, VV has increased sensitivity to certain phenological stages as the incidence angle increases. Additionally, a periodic oscillation on a sub-daily scale is observed when creating coherence time series with increasing temporal baseline. Factors which strongly contribute to these oscillations are the daily cycles of temperature, soil moisture and vegetation water dynamics.

Welcome to the proceedings of the "Workshop of Physical Agents", a distinguished gathering that brings together researchers and practitioners from diverse fields such as computer vision, robotics, cognitive architectures, and machine learning. This workshop serves as a vibrant platform for the exchange of innovative ideas, cutting-edge research, and transformative technologies that are shaping the future of physical agents. Physical agents, encompassing autonomous robots, intelligent systems, and adaptive software agents, are becoming increasingly integral to various aspects of our lives. From enhancing industrial automation to advancing medical diagnostics, these agents are pushing the boundaries of what is possible in both the digital and physical realms. The interdisciplinary nature of this workshop reflects the multifaceted challenges and opportunities in developing, deploying, and understanding these sophisticated systems. This collection of scientific papers represents the latest advancements in the field. Our contributors delve into a wide array of topics, showcasing the breadth and depth of current research. The papers are organized into several key themes: Computer Vision, Intelligent Robotics, Machine learning, Social Robotics, Intelligent Robotics and Cognitive robotics. The diversity of topics covered in this workshop underscores the collaborative spirit and interdisciplinary approach essential for advancing the field of physical agents. Each paper has undergone a rigorous peer-review process, ensuring the highest standards of scientific quality and relevance. We extend our heartfelt gratitude to the authors for their valuable contributions, the reviewers for their insightful feedback, and the organizing committee for their dedication and hard work in making this workshop a success. We also thank our sponsors and supporters for their generous contributions, which have been instrumental in facilitating this event. As you explore these proceedings, we hope

In this study, a general three-component polarimetric SAR interferometry (PolInSAR) target decomposition framework is proposed by modifying the existing generalized surface, double-bounce, and volume models. The resulting general models are then compared to the original Freeman-Durden modeling strategy. Three types of generalized volume scattering models (generalized volume scattering model (GVSM), simplified Neumann volume scattering model (SNVSM) and simplified adaptive volume scattering model (SAVSM)) were employed. Simulated L-band PolInSAR data over deciduous and pine forest stands generated by PolSARpro and DLR P-band airborne PolInSAR data over a tropical forest area from the AfriSAR 2016 campaign were used for performance analysis. A qualitative comparison of the decomposition results shows that the three generalized volume scattering models generally deviate from the Freeman-Durden model, showing that the GVSM and SNVSM models have very similar results. In the case of airborne data over tropical forests, a tomographic synthetic aperture radar (TomoSAR) profile was also computed and used as a benchmark for comparison with the phase-center profiles of all four volume-scattering components. Not only do the GVSM and SNVSM models exhibit similar results between them (as with simulated data), but also a better match with the HV TomoSAR profile.

La traducción especializada desempeña un papel crucial en los ámbitos profesionales. Una óptima comunicación interlingüística es necesaria para la comprensión de la realidad económica y financiera en distintos países. En este proyecto, se plantea la duda de si se cumple en la práctica el principio de biunivocidad y monorreferencialidad propio de los términos especializados. Se analiza una serie de términos con un alto nivel de abstracción y especialización, y que se emplean en finanzas. Para ello, se compila un corpus de textos de distintas entidades financieras en español e inglés, de forma que sea posible comparar el mismo concepto in vivo en ambos idiomas y en el mismo contexto. De esta forma, se comprueba si en todas las entidades financieras se ha utilizado el mismo término o si existe variación. Como en varios términos se produce variación denominativa o semántica, se analizan en profundidad para diferenciar las causas., Specialized translation plays a crucial role in professional applications. Optimal interlingual communication is necessary to understand the economic and financial reality in different countries. This paper aims to call into question the accomplishment of the univocity principle in specialized texts. A set of abstract and specialized financial terms are examined. With the purpose of extracting these terms, a bilingual corpus of texts issued by several financial institutions is compiled, so that it is possible to compare the use of a concept in vivo in both English and Spanish, given the same context. Thus, we test whether all the financial institutions use the same term or there actually is variation. The results show that many terms do have denominative or semantic variation. Finally, some reasons of variation that apply in our corpus are analyzed.

The production of wild fish has remained relatively stable in the last two decades, whereas aquaculture organism production has increased to the point where it has exceeded wild catches. In this context, accurate and up-to-date information about the current usage of marine areas for aquaculture is crucial for the planning of marine activities. However, this data is often limited to national authorities, and discrepancies between planned and real practices can arise in available data. In this study, a novel methodology to automatically map and verify the current activity of aquaculture crops across European regions based on freely available satellite data is proposed. The European Space Agency’s (ESA) Sentinel-1 mission provides Synthetic Aperture Radar (SAR) images, which serve as the basis for the analysis. Multiple SAR images of the same locations are processed using ESA Sentinel Application Platform (SNAP) software and merged to remove temporal noise-like artifacts caused by factors such as ships and waves. Next, the iDPolRAD algorithm is employed to detect potential aquaculture sites, which initially include noise from coastal zones and unwanted human and natural structures that pass through the filter. The aquaculture sites are classified using a ResNet18 model with 93% of the sites correctly classified. This implies that it is feasible to monitor marine areas using satellite radar data to track aquaculture areas. However, generalization power across regions is poor likely due to the diversity of types of structures used and species cultivated. Further studies are needed to investigate factors influencing the detectability of different aquaculture sites such as cage geometry or SAR image resolution in order to enhance the accuracy and comprehensiveness of the mapping process. This study highlights the potential of SAR data, coupled with image processing and classification techniques, as a viable means to map large marine areas dedicated to aquaculture.

Recent advances in generative models have skyrocketed the popularity of conversational chatbots and have revolutionized the way we interact with artificial intelligence. At the same time, research has shown that machine learning models can unconsciously reflect and amplify human biases. This is particularly dangerous for generative models given the huge popularity of such technologies. Specifically, a fundamental source of bias of such technologies is the resources on which the models are trained. To address this issue, this paper proposes a methodology to analyze intrinsic gender bias in Natural Language Generation (NLG). Some works already propose metrics and approaches to measure bias in the Natural Language processing field. However, there is a lack of standard methodology to measure gender bias in NLG. Therefore, adapting the Bias Score approach, our proposal involves three sequential stages applied to individual texts to detect intrinsic gender bias on NLG effectively. Those steps are as follows: (i) word scoring; (ii) word filtering; and (iii) generative-word analysis. This methodology is applied to recent datasets and pre-trained models widely used for the generation of text with common sense. In particular, this paper analyzes the potential gender bias in CommonGen and C2 Gen datasets and the SimpleNLG and T5 models. The results show the ability of the proposed methodology to detect gender bias in word distributions, presenting a strong correlation with the words typically associated with a specific gender. Results indicate that both tested datasets are intrinsically gender-biased, and therefore, tested models fine-tuned with those datasets also are.

A variety of methods for 3D object recognition and registration based on a deep learning pipeline have recently emerged. Nonetheless, these methods require large amounts of data that are not easy to obtain, sometimes rendering them virtually useless in real-life scenarios due to a lack of generalization capabilities. To counter this, we propose a novel local descriptor that takes advantage of the fractal dimension. For each 3D point, we create a descriptor by computing the fractal dimension of the neighbors at different radii. Our redmethod has many benefits, such as being agnostic to the sensor of choice and noise, up to a level, and having few parameters to tinker with. Furthermore, it requires no training and does not rely on semantic information. We test our descriptor using well-known datasets and it largely outperforms Fast Point Feature Histogram, which is the state-of-the-art descriptor for 3D data. We also apply our descriptor to a registration pipeline and achieve accurate three-dimensional representations of the scenes, which are captured with a commercial sensor.

The demonstrated efficacy of interferometric synthetic aperture radar (InSAR) techniques has spurred the development of innovative SAR satellite missions like BIOMASS and NiSAR, poised for extensive application in forest monitoring. Nevertheless, prevailing methodologies for retrieving forest variables, including forest height and above-ground biomass, encounter substantial limitations. Traditionally, successful forest mapping necessitates a non-zero spatial perpendicular baseline, full polarimetry, and a relatively small (close-to-zero) temporal baseline. This study presents a novel approach for extracting forest biophysical variables by modeling the temporal decorrelation of repeat-pass InSAR coherence. We explore a hypothesis regarding the potential relationship between the temporal decorrelation of InSAR coherence and forest variables, such as tree height and aboveground biomass. This hypothesis is tested across diverse test sites in Finland, Canada, and Germany. Our findings suggest a viable means of extracting forestry information by quantifying the temporal decorrelation of C-Band InSAR coherence. We establish a clear connection between the temporal decay rate and crucial forest variables, such as forest above-ground biomass and tree height.

The incorporation of machine learning algorithms into high-risk decision-making tasks has raised some alarms in the scientific community. Research shows that machine learning-based technologies can contain biases that cause unfair decisions for certain population groups. The fundamental danger of ignoring this problem is that machine learning methods can not only reflect the biases present in our society but could also amplify them. This article presents the design and validation of a technology to assist the fair automation of classification problems. In essence, the proposal is based on taking advantage of the intermediate solutions generated during the resolution of classification problems through using Auto-ML tools, in particular, AutoGOAL, to create unbiased/fair classifiers. The technology employs a multi-objective optimization search to find the collection of models with the best trade-offs between performance and fairness. To solve the optimization problem, we introduce a combination of Probabilistic Grammatical Evolution Search and NSGA-II. The technology was evaluated using the Adult dataset from the UCI repository, a common benchmark in related research. Results were compared with other published results in scenarios with single and multiple fairness definitions. Our experiments demonstrate the technology's ability to automate classification tasks while incorporating fairness constraints. Additionally, our method achieves competitive results against other bias mitigation techniques. A notable advantage of our approach is its minimal requirement for machine learning expertise, thanks to its Auto-ML foundation. This makes the technology accessible and valuable for advancing fairness in machine learning applications. The source code is available online for the research community.

The massive use of social media has increased the ease of dissemination of information. Unfortunately, every type of information can be massively disseminated (even deceptive information deliberately created to mislead). In this research we introduce a deep analysis of linguistic cues (i.e., adjectives, pronouns, complex syntax, emotion words, etc.) that can lead to distinguish which texts can be deceptive. The research focuses on extracting a combination of features: content-based, context-based, readability, virality and information richness. The objective is to test if current NLP tools for deception detection can extract in a satisfactory way these features and to examine the grade of explainability that these systems offer. The methodology starts from a multidisciplinary point of view, focusing in elaborate an integrative research. To pursue the objective of this research, we also combine both analytical and empirical methodologies. The expected impact is to enhance the ability to distinguish misinformation by improving the accuracy and transparency of deception detection systems for everyone.

Data obtained during a ground-based SAR experiment and an associated field campaign have been exploited to study the rate and sources of decorrelation in an agricultural test site in the conditions of observation of a geosynchronous SAR. It was found that the scene is less affected by temporal decorrelation when the primary image is acquired during night time or early morning. Additionally, a periodic oscillation on a sub-daily scale was observed when creating coherence time series with increasing temporal baseline. Two factors which strongly contribute to these oscillations are the daily cycles of soil moisture and evapotranspiration.

Acquiring high-resolution and accurate forest height is essential for estimating terrestrial carbon storage and detecting changes. Since 2010, TanDEM-X has obtained an unprecedented global interferometric SAR dataset and has been widely used to estimate forest height. However, its accuracy is limited due to insufficient observational information. Since September 2018, NASA’s Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) has acquired global discrete surface elevation and forest canopy height, providing an excellent opportunity to improve the performance of TanDEM-X forest height estimation. Based on our previous research, this paper proposes some methods to estimate highresolution and large-scale forest height using these two missions combined with optical remote sensing data, and estimates forest height in Spain. As validated against LiDAR data, the RMSE of the estimated optimal average forest height ranges from 1.83 m to 3.70 m, at the resolution of onehectare forest stands.

TanDEM-X InSAR data has exhibited commendable performance in forest height inversion, with the semi-empirical SINC model (SeEm-SINC) proving its robustness in flat regions. However, the relationship between InSAR coherence and forest height become unreliable in forest scenes of mountainous terrain. Significant overestimation or underestimation of forest height occurs when using coherence only in areas with positive or negative slope, and the bias is strongly correlated with the range slope. Building upon the SeEm-SINC model, we investigate the relationship between slope and the bias involved in forest height inversion. Additionally, we present a forest height inversion approach utilizing TanDEM-X InSAR ascending and descending data. Experimental results demonstrate that this method effectively mitigates estimation deviations caused by slope in forested mountainous areas.

In this paper, we explore the utilisation of the complex coherence between HH and VV channels as a tool for crop monitoring. The analysis includes the examination of time series of polarimetric data collected from both airborne and space-borne platforms at L and C band. Results suggest that the L band exhibits a higher sensitivity to crop changes than the C-band in both amplitude and phase of the coherence between the co-polar channels. Moreover, incidence angle is identified as a factor influencing the values of the complex coherence, and hence its sensitivity to crop features.

Large Language Models (LLMs) are trained on extensive data, and their performance and behaviour are shaped by the quality and diversity of this data. However, texts used for training these Artificial Intelligence (AI) systems often fail to reflect the wide range of human experiences and identities, which can result in reflecting and amplifying biases. Consequently, if a LLM is trained on biased texts, it can magnify biases when generating text. Regarding the continuous generation of information, it complicates the understanding of such texts, specially to users with cognitive impairment. This issue has made it necessary to resort to automated processes to make information available to all users. Therefore, Automatic Text Simplification (ATS) arises with the aim of approaching the challenge of automatically transforming an original text into a simplified and easier one. More specifically, this task will focus on the medical domain, to make medical texts accessible to all society.

A series of X-band images acquired by TerraSAR-X, TanDEM-X and PAZ in the same orbit configuration over an agricultural area has been exploited to assess the potential of repeat-pass coherence for crop monitoring. Temporal baselines of different length are considered. The shorter temporal baselines are less affected by temporal decorrelation outside the growing season. Then, depending on the physical features of the plants, coherence decreases to a minimum level at the middle of the cultivation cycle, increasing again after harvest. The minimum coherence depends on the temporal baseline for short crops, whereas it is the same for tall ones.

TanDEM-X interferometric synthetic aperture radar (InSAR) data have demonstrated promising advantages and potential in recent years for the inversion of forest height. InSAR coherence becomes the primary input feature when a precise digital terrain model (DTM) is unavailable, but the relationship between InSAR coherence and forest height remains uncertain because of the complexity of forest scenes. In this paper, a method for retrieving canopy height in Mediterranean forests, characterised by short and sparse trees, using a single-pass bistatic TanDEM-X InSAR dataset is proposed. To improve the accuracy of forest height inversion from the uncertain correlation between InSAR coherence and canopy height, we begin by using the established SINC model with two semi-empirical parameters and then expand the single curve into a collection of three curves, forming the Multi-SINC model. To determine the optimal relationship (curve) between TanDEM-X InSAR coherence and canopy height, the problem is shifted from parameter inversion to classification. To solve the problem, we used optical remote sensing data, a small amount of LiDAR data, and TanDEM-X InSAR data in combination with machine learning for classification. As a proof-of-concept, we conducted forest height retrieval at two study sites in Spain with complex terrain and diverse forest types. The results were verified by comparing them with LiDAR product forest height, which demonstrated improved performance (RMSE = 2.49 m and 1.7 m) compared to the SeEm-SINC model (RMSE = 3.28 m and 2.36 m).

The mapping and classification of active landslides in high mountainous regions provide crucial information about the location and types of geohazards. Additionally, this process plays a vital role in ensuring the safety of the geological environments in mountainous towns. In this study, we presented a refined InSAR approach for mapping and classifying active landslide hazards in Deqin County, Tibetan Plateau, China. The study area is characterized by a high altitude and extremely rugged terrain. Consequently, conventional InSAR methods are limited in precisely estimating landslide deformation owing to severe atmospheric delays. To this end, we first propose a block-based linear model to correct tropospheric artifacts. This model considers the spatial variability of the atmosphere and provides an opportunity to accurately estimate heterogeneous atmospheric delays over high mountainous areas without any external data. Compared with the traditional global-window linear model and the GACOS approach, the new method demonstrated outstanding performance in reducing atmospheric artifacts. Second, based on the knowledge mapping of landslide types, we proposed a semi-automatic procedure to map and classify landslides using InSAR-derived displacements and auxiliary data (i.e., C-index and high resolution optical images). Our results obtained from ascending and escending Sentinel-1 images revealed, for the first time, that there were 317 active landslides in Deqin County between May 2017 and June 2021. Among these, 10.7% were associated with slide activity, 7.9% with fall deformation, and the majority (81.4%) with flow movement. These results were cross-verified and evaluated using an a priori inventory map obtained from the visual interpretation of optical images and geological field surveys. This study demonstrates that InSAR can accurately map and classify active landslides over difficult mountainous terrains, provided the associated phase errors are effectively restrained.

The photovoltaic (PV) industry is seeing a significant shift toward large-scale solar plants, where traditional inspection methods have proven to be time-consuming and costly. Currently, the predominant approach to PV inspection using unmanned aerial vehicles (UAVs) is based on the capture and detailed analysis of aerial images (photogrammetry). However, the photogrammetry approach presents limitations, such as an increased amount of useless data and potential issues related to image resolution that negatively impact the detection process during high-altitude flights. In this work, we develop a visual servoing control system with dynamic compensation using nonlinear model predictive control (NMPC) applied to a UAV. This system is capable of accurately tracking the middle of the underlying PV array at various frontal velocities and height constraints, ensuring the acquisition of detailed images during low-altitude flights. The visual servoing controller is based on extracting features using RGB-D images and employing a Kalman filter to estimate the edges of the PV arrays. Furthermore, this work demonstrates the proposal in both simulated and real-world environments using the commercial aerial vehicle (DJI Matrice 100), with the purpose of showcasing the results of the architecture.

The spread of fake news (FN) has attracted attention from disciplines ranging from social sciences to Artificial Intelligence. This work is novel because it explores the news-sharing behaviour of social-media users, focussing on those that spread FN, rather than the psychological motivations behind them. The 14-item Risky News-Sharing Quotient (RNSQ) was developed and Exploratory Factor Analysis discovered three relevant factors: (i) news-sharing behaviour that contributes to debunking FN; (ii) news-sharing frequency and attitudes to sharing; and (iii) news-sharing behaviour that contributes to the spread of FN. The study, conducted among university students, found that 75% reported risky news-sharing behaviour that spreads FN. No link was found between perceiving FN as a problem and debunking it. Moreover, 83% of survey participants were unable to identify a FN story. Overall, the findings suggest an inability to apply knowledge of the relevant FN detection strategies to debunk FN, but importantly an apparent lack of motivation to check the veracity of a news story. From these conclusions, better-informed educational intervention strategies can be implemented to address the FN problem in-situ, such as promoting the importance of responsible news-sharing by raising awareness of how the spread of FN can impede the proper functioning of societies.

Recent research has shown that the unaware automation of high-risk decision-making tasks can result in unfair decisions being made. The most common approaches to address this problem adopt definitions of fairness based on protected attributes. Precise annotation of protected attributes enables the application of bias mitigation techniques to commonly unlabeled kinds of data (e.g., images, text, etc.). This paper proposes a framework to automatically annotate protected attributes in data collections. The framework focuses on providing a single interface to annotate protected attributes of different types (e.g., gender, race, etc.) and from different kinds of data. Internally, the framework coordinates multiple sensors to produce the final annotation. Several sensors for textual data are proposed. An optimization search technique is designed to tune the framework to specific domains. Additionally, a small dataset of movie reviews —annotated with gender and sentiment— was created. The evaluation in datasets of texts from diverse domains shows the quality of the annotations and their effectiveness to be used as a proxy to estimate fairness in datasets and machine learning models. The source code is available online for the research community.

The new global scenario imposed by the SARS-CoV-2 virus has given rise to an atypical and problematic situation in multiple spheres. In very little time, the abrupt change from face-to-face to remote has not only required a rapid widespread use of digital technology, but also a change in methodology and communicative interactions. In the field of education, teachers have had to interact in new environments, with the combined use of face-to-face and non-face-to-face teaching being a major challenge. This paper presents the design and implementation of a cyber presence system for educational environments using Microsoft’s HoloLens 2 Mixed Reality (MR) headset. A software tool is developed that improves teaching scenarios through communication in mixed environments. The tool enables teachers to integrate the students in the classroom in a common space with remote students connected by videoconference. Our system is not limited to education, however, as it can also be deployed in any setting that requires remote communication, such as companies and governmental institutions.

Hoy en día, las pequeñas o medianas empresas se encuentran en una situación complicada a la hora de dar el salto hacia la robotización de sus procesos industriales, ya sea por motivos económicos o por desconocimiento de estas tecnologías. Es por eso que este sistema robotizado representa un avance significativo en la automatización industrial en un sector muy tradicional y manufacturero, el sector del juguete. Esta tecnología ofrecerá una solución eficiente, precisa y flexible al proceso de producción de juguetes, concretamente en el proceso de ensamblaje de la muñeca, mediante la optimización de recursos, lo que deriva en una reducción directa de costes y en una mayor adaptabilidad a los cambios en la demanda del mercado. Por último, el operario no será reemplazado en el proceso completo, únicamente en las tareas que requieran grandes esfuerzos o sean tediosas. La complejidad añadida a este proceso de ensamblaje es la manipulabilidad del producto, ya que tratamos con un plástico flexible y deformable., Nowadays, small and medium-sized companies find themselves in a complicated situation when it comes to making the leap towards the robotisation of their industrial processes, either for economic reasons or because they are unfamiliar with these technologies. That is why this robotic system represents a significant advance in industrial automation in a very traditional and manufacturing sector, the toy sector. This technology will offer an efficient, precise and flexible solution to the toy production process, specifically in the toy doll assembly process, by optimising resources, resulting in a direct reduction in costs and greater adaptability to changes in market demand. Finally, the operator will not be replaced in the entire process, only in the tasks that require great effort or are tedious. The added complexity to this assembly process is the manipulability of the product, because it is a flexible and deformable plastic.

Monitoring crop phenology is essential for managing field disasters, protecting the environment, and making decisions about agricultural productivity. Because of its high timeliness, high resolution, great penetration, and sensitivity to specific structural elements, synthetic aperture radar (SAR) is a valuable technique for crop phenology estimation. Particle filtering (PF) belongs to the family of dynamical approach and has the ability to predict crop phenology with SAR data in real time. The observation equation is a key factor affecting the accuracy of particle filtering estimation and depends on fitting. Compared to the common polynomial fitting (POLY), machine learning methods can automatically learn features and handle complex data structures, offering greater flexibility and generalization capabilities. Therefore, incorporating two ensemble learning algorithms consisting of support vector machine regression (SVR), random forest regression (RFR), respectively, we proposed two machine learning-aided particle filtering approaches (PF-SVR, PF-RFR) to estimate crop phenology. One year of time-series Sentinel-1 GRD SAR data in 2017 covering rice fields in Sevilla region in Spain was used for establishing the observation and prediction equations, and the other year of data in 2018 was used for validating the prediction accuracy of PF methods. Four polarization features (VV, VH, VH/VV and Radar Vegetation Index (RVI)) were exploited as the observations in modeling. Experimental results reveals that the machine learning-aided methods are superior than the PF-POLY method. The PF-SVR exhibited better performance than the PF-RFR and PF-POLY methods. The optimal outcome from PF-SVR yielded a root-mean-square error (RMSE) of 7.79, compared to 7.94 for PF-RFR and 9.1 for PF-POLY. Moreover, the results suggest that the RVI is generally more sensitive than other features to crop phenology and the performance of polarization features presented consistent among all methods, i.

Medical image datasets are essential for training models used in computer-aided diagnosis, treatment planning, and medical research. However, some challenges are associated with these datasets, including variability in data distribution, data scarcity, and transfer learning issues when using models pre-trained from generic images. This work studies the effect of these challenges at the intra- and inter-domain level in few-shot learning scenarios with severe data imbalance. For this, we propose a methodology based on Siamese neural networks in which a series of techniques are integrated to mitigate the effects of data scarcity and distribution imbalance. Specifically, different initialization and data augmentation methods are analyzed, and four adaptations to Siamese networks of solutions to deal with imbalanced data are introduced, including data balancing and weighted loss, both separately and combined, and with a different balance of pairing ratios. Moreover, we also assess the inference process considering four classifiers, namely Histogram, kNN, SVM, and Random Forest. Evaluation is performed on three chest X-ray datasets with annotated cases of both positive and negative COVID-19 diagnoses. The accuracy of each technique proposed for the Siamese architecture is analyzed separately. The results are compared to those obtained using equivalent methods on a state-of-the-art CNN, achieving an average F1 improvement of up to 3.6%, and up to 5.6% of F1 for intra-domain cases. We conclude that the introduced techniques offer promising improvements over the baseline in almost all cases and that the technique selection may vary depending on the amount of data available and the level of imbalance.

Current climate and weather changes urged the scientific community to find ways to observe and understand the relatively rapid land-atmosphere processes related to the water cycle. The HydroSoil project, funded by the European Space Agency, was proposed to experimentally assess the use of Synthetic Aperture Radar in future short revisit time missions based on Geosynchronous Orbital platforms. This paper describes the HydroSoil facility providing continuous monitoring of an agricultural field with a 10-minute temporal resolution and a spatial resolution in the order of 1m2. Gathered data from the full Pol Ground-Based SAR are being used to investigate the capability of Geosynchronous SAR missions to retrieve soil moisture and to measure vegetation parameters. To achieve the project's objectives, intensive procurement of ground truth data encompassing continuous acquisition of relevant meteorological parameters like air humidity, temperature, precipitation wind speed and direction, pressure, and solar irradiation together with other vegetation characteristics under study has been mandatory. The different densities, sizes, and water content of two different crops, barley and corn, together with the data generated, will enable the assessment of the missions' ability to extract essential scene characteristics, such as soil moisture and crop water content, for use in future geostationary missions.

Optimal robotic grasping cannot be limited to the estimation of object grasping pose using vision-based methods. It is necessary to use tactile sensors to learn the physical properties of the objects that are to be grasped. In this work, we integrated two Contactile force-based tactile sensors with a 2F-140 ROBOTIQ gripper and a UR5 robot to estimate the volume of a waterf illed container using Multilayer Perceptron (MLP) neural networks. During experimentation, we trained and evaluated different MLPs varying the input forces (Fx, Fy, Fz) in a task of discrete-volume regression in a range of between 0ml and 300ml. The preliminary proposed approach is compared with an algebraic method based on the diagram of the equilibrium of forces, proving that our results are more precise, obtaining a R2 value of 8% higher in the worst-case scenario, and of 30% in the best., Un agarre robótico óptimo no puede limitarse a la estimación de pose de agarre del objeto mediante visión. Se hace necesario el uso de sensores táctiles para conocer las propiedades físicas de los objetos que se agarran. En este trabajo, integramos dos sensores táctiles Contactile basados en la fuerza con una pinza ROBOTIQ 2F-140 y un robot UR5, para estimar el volumen de un recipiente lleno de agua utilizando redes neuronales Perceptrón Multicapa (MLP). Durante la experimentación entrenamos y evaluamos diferentes MLPs variando las fuerzas de entrada (Fx, Fy, Fz) en una tarea de regresión de volumen discreto en un rango de entre 0ml y 300ml. El enfoque preliminar propuesto se compara con un método algebraico basado en el diagrama del equilibrio de fuerzas, demostrando que nuestros resultados son más precisos, obteniendo un valor R2 un 8% superior en el peor de los casos, y del 30% en el mejor.

Feature-based geo-localization relies on associating features extracted from aerial imagery with those detected by the vehicle's sensors. This requires that the type of landmarks must be observable from both sources. This lack of variety of feature types generates poor representations that lead to outliers and deviations produced by ambiguities and lack of detections, respectively. To mitigate these drawbacks, in this letter, we present a dynamically weighted factor graph model for the vehicle's trajectory estimation. The weight adjustment in this implementation depends on information quantification in the detections performed using a LiDAR sensor. Also, a prior (GNSS-based) error estimation is included in the model. Then, when the representation becomes ambiguous or sparse, the weights are dynamically adjusted to rely on the corrected prior trajectory, mitigating outliers and deviations in this way. We compare our method against state-of-the-art geo-localization ones in a challenging and ambiguous environment, where we also cause detection losses. We demonstrate mitigation of the mentioned drawbacks where the other methods fail.

To address the challenge of retrieving sub-canopy topography using single-baseline single-polarization TanDEM-X InSAR data, we propose a novel InSAR processing framework. Our methodology begins by employing the SINC model to estimate the penetration depth (PD). Subsequently, we establish a linear relationship between PD and phase center height (PCH) to generate a wall-to-wall PCH product. To achieve this, space-borne LiDAR data are employed to capture the elevation bias between actual ground elevation and InSAR-derived elevation. Finally, the sub-canopy topography is derived by subtracting the PCH from the conventional InSAR-based DEM. Moreover, this approach enables the simultaneous estimation of forest height from single-baseline TanDEM-X data by combining the estimated PD and PCH components. The approach has been validated against Airborne Lidar Scanning data over four diverse sites encompassing different forest types, terrain conditions, and climates. The derived sub-canopy topography in the boreal and hemi-boreal forest sites (Krycklan and Remningstorp) demonstrated notable improvement in accuracy. Additionally, the winter acquisitions outperformed the summer ones in terms of inversion accuracy. The achieved RMSEs for the winter scenarios were 2.45 m and 3.83 m, respectively, representing a 50% improvement over the InSAR-based DEMs. And the forest heights are also close to the ALS measurements, with RMSEs of 2.70 m and 3.33 m, respectively. For the Yanguas site in Spain, characterized by rugged terrain, sub-canopy topography in forest areas was estimated with an accuracy of 4.27m, a 35% improvement over the original DEM. For the denser tropical forest site, only an average elevation bias could be corrected.

Jump height tests are employed to measure lower-limb muscle power of athletic and non-athletic populations. The most popular instruments for this purpose are jump mats and, in recent years, smartphone apps, which compute jump height through the manual annotation of video recordings and recently automatically using the sound produced during the jump to extract the flight time. In a previous work, the afore-mentioned sound systems were presented by the authors in which the take-off and landing events from the audio recordings of jump executions were obtained using classical signal processing. In this work, a more precise, noise-immune, and robust system, capable of working in the most unfavorable environments, is presented. The system uses a deep neural network trained specifically for this purpose. More than 300 jumps were recorded to train and validate the network performance. The ground truth was a jump mat, providing a slightly better accuracy in quiet and medium quiet environments but excellent accuracy in noisy and complicated ones. The developed audio-based system is a trustworthy instrument for measuring jump height accurately in any kind of environment, providing a perfect measurement tool that can be accessed through a mobile phone in the form of an app.

Tourism represents a crucial socio-economic pillar globally, yet the multifaceted challenges it poses necessitate innovative management approaches. The paradigm of smart tourism harnesses advanced data analytics tools to promote both profitability and sustainability in tourist destinations, leading to new levels of destination smartness. Accurate tourist occupancy prediction, particularly in areas dominated by second-home accommodations where traditional hospitality data may be insufficient, plays a key role in optimising tourism management. To address this data gap, our prior research employed ARIMA modelling on Airbnb booking time series and analysed tourism-related Twitter conversations to forecast occupancy levels in Torrevieja (Alicante); a prominent second-home tourism destination in Southeastern Spain. In this extended study, we delve deeper into the realm of social sensing by utilising bidirectional encoder representations from transformers (BERT) for topic modelling. Our methodology involves the processing and analysis of Twitter data to identify prominent themes related to Torrevieja. The findings not only reveal nuanced perceptions and discussions about the destination but also underscore the effectiveness of BERT in capturing intricate topic dynamics. Importantly, this work highlights how the alignment of specific topics with booking patterns can further enhance predictive accuracy for tourist occupancy, presenting a robust toolkit for stakeholders in the tourism sector.

Classifying logo images is a challenging task as they contain elements such as text or shapes that can represent anything from known objects to abstract shapes. While the current state of the art for logo classification addresses the problem as a multi-class task focusing on a single characteristic, logos can have several simultaneous labels, such as different colours. This work proposes a method that allows visually similar logos to be classified and searched from a set of data according to their shape, colour, commercial sector, semantics, general characteristics, or a combination of features selected by the user. Unlike previous approaches, the proposal employs a series of multi-label deep neural networks specialized in specific attributes and combines the obtained features to perform the similarity search. To delve into the classification system, different existing logo topologies are compared and some of their problems are analysed, such as the incomplete labelling that trademark registration databases usually contain. The proposal is evaluated considering 76,000 logos (seven times more than previous approaches) from the European Union Trademarks dataset, which is organized hierarchically using the Vienna ontology. Overall, experimentation attains reliable quantitative and qualitative results, reducing the normalized average rank error of the state-of-the-art from 0.040 to 0.018 for the Trademark Image Retrieval task. Finally, given that the semantics of logos can often be subjective, graphic design students and professionals were surveyed. Results show that the proposed methodology provides better labelling than a human expert operator, improving the label ranking average precision from 0.53 to 0.68.

Este trabajo se propone demostrar que el concepto de la menopausia está asociado a connotaciones negativas. Para ello, nos centramos en el análisis de sentimientos del léxico extraído de una encuesta y en el uso de recursos de procesamiento de lenguaje natural: EmoLex y BabelNet. Tras el análisis de sentimientos y la consulta de definiciones en diferentes fuentes, se han hecho propuestas de redefinición de la terminología relevante relacionada con la menopausia desde una perspectiva de género. Este estudio se enmarca en el proyecto digitender, cuyo propósito es la creación de recursos terminológicos sobre la salud de la mujer., The aim of this article is to show that the concept of menopause is associated with negative connotations. To this end, we performed a sentiment analysis on the lexicon extracted from a survey on menopause and used the Natural Language Processing tools EmoLex and BabelNet. Following the sentiment analysis and consultation of definitions in different sources, suggestions are made as to how to redefine relevant terminology related to menopause from a gender perspective. This study is part of the digitender project, which aims to create terminological resources on the topic of women’s health.

The attention level of students in a classroom can be improved through the use of Artificial Intelligence (AI) techniques. By automatically identifying the attention level, teachers can employ strategies to regain students' focus. This can be achieved through various sources of information. One source is to analyze the emotions reflected on students' faces. AI can detect emotions, such as neutral, disgust, surprise, sadness, fear, happiness, and anger. Additionally, the direction of the students' gaze can also potentially indicate their level of attention. Another source is to observe the students' body posture. By using cameras and deep learning techniques, posture can be analyzed to determine the level of attention. For example, students who are slouching or resting their heads on their desks may have a lower level of attention. Smartwatches distributed to the students can provide biometric and other data, including heart rate and inertial measurements, which can also be used as indicators of attention. By combining these sources of information, an AI system can be trained to identify the level of attention in the classroom. However, integrating the different types of data poses a challenge that requires creating a labeled dataset. Expert input and existing studies are consulted for accurate labeling. In this paper, we propose the integration of such measurements and the creation of a dataset and a potential attention classifier. To provide feedback to the teacher, we explore various methods, such as smartwatches or direct computers. Once the teacher becomes aware of attention issues, they can adjust their teaching approach to re-engage and motivate the students. In summary, AI techniques can automatically identify the students' attention level by analyzing their emotions, gaze direction, body posture, and biometric data. This information can assist teachers in optimizing the teaching-learning process.

When carrying out robotic manipulation tasks, objects occasionally fall as a result of the rotation caused by slippage. This can be prevented by obtaining tactile information that provides better knowledge on the physical properties of the grasping. In this paper, we estimate the rotation angle of a grasped object when slippage occurs. We implement a system made up of a neural network with which to segment the contact region and an algorithm with which to estimate the rotated angle of that region. This method is applied to DIGIT tactile sensors. Our system has additionally been trained and tested with our publicly available dataset which is, to the best of our knowledge, the first dataset related to tactile segmentation from non-synthetic images to appear in the literature, and with which we have attained results of 95% and 90% as regards Dice and IoU metrics in the worst scenario. Moreover, we have obtained a maximum error of ≈ 3° when testing with objects not previously seen by our system in 45 different lifts. This, therefore, proved that our approach is able to detect the slippage movement, thus providing a possible reaction that will prevent the object from falling.

The recognition of patterns that have a time dependency is common in areas like speech recognition or natural language processing. The equivalent situation in image analysis is present in tasks like text or video recognition. Recently, Convolutional Recurrent Neural Networks (CRNN) have been broadly applied to solve these tasks in an end-to-end fashion with successful performance. However, its application to Optical Music Recognition (OMR) is not so straightforward due to the presence of different elements sharing the same horizontal position, disrupting the linear flow of the timeline. In this paper, we study the ability of the state-of-the-art CRNN approach to learn codes that represent this disruption in homophonic scores. In our experiments, we study the lower bounds in the recognition task of real scores when the models are trained with synthetic data. Two relevant conclusions are drawn: (1) Our serialized ways of encoding the music content are appropriate for CRNN-based OMR; (2) the learning process is possible with synthetic data, but there exists a glass ceiling when recognizing real sheet music.

End-to-end solutions have brought about significant advances in the field of Optical Music Recognition. These approaches directly provide the symbolic representation of a given image of a musical score. Despite this, several documents, such as pianoform musical scores, cannot yet benefit from these solutions since their structural complexity does not allow their effective transcription. This paper presents a neural method whose objective is to transcribe these musical scores in an end-to-end fashion. We also introduce the GrandStaff dataset, which contains 53,882 single-system piano scores in common western modern notation. The sources are encoded in both a standard digital music representation and its adaptation for current transcription technologies. The method proposed in this paper is trained and evaluated using this dataset. The results show that the approach presented is, for the first time, able to effectively transcribe pianoform notation in an end-to-end manner.

Hoy en día, existen muchas tareas repetitivas y tediosas llevadas a cabo por personas que pueden ser reemplazadas por sistemas inteligentes robóticos, permitiendo a esos operarios realizar otro tipo de trabajos que requieran más destreza. La tarea abordada es el ensamblaje automático de muñecos de juguete en el entorno industrial. El reto principal es el material deformable del cual están hechos los muñecos, cuyo comportamiento físico es dinámico, ya que sus características cambian dependiendo de la fuerza aplicada y la temperatura a la que está sometido. Nuestra propuesta es una comparación de diferentes métodos para realizar la tarea, centrándonos en la manipulación de objetos deformables. Por un lado, el método propuesto capta la información del proceso realizado previamente por un operario experto, permitiendo grabar los datos necesarios (movimientos, velocidades, posiciones, etc). Por otro lado, un módulo de percepción se ha llevado a cabo utilizando algoritmos basados en visión para detectar las piezas y realizar el ensamblaje mediante un robot manipulador., Nowadays, there are a lot of repetitive and tedious tasks carried out by people who can be replaced by an intelligent robotic system, allowing the operators to perform other kind of dexterous works. The tackled task is the automatic assembling of a toy doll in a real environment. The main challenge is the soft material the dolls are made of, whose physical behaviour dynamic, as their features change depending of the applied force and temperature. Our proposal is a comparison of different approaches to perform the task, focusing on the handling of such flexible materials. On the one hand, the proposed method acquires the information of the process by the previous demonstration of an expert operator, allowing to record all necessary data (movements, positions, velocities, etc). On the other hand, a perception module is developed employing vision-based algorithms to detect the pieces and to perform the assembly using a robot manipulator.

Around a third of the total population of Europe suffers from mental disorders. The use of electroencephalography (EEG) together with Machine Learning (ML) algorithms to diagnose mental disorders has recently been shown to be a prominent research area, as exposed by several reviews focused on the field. Nevertheless, previous to the application of ML algorithms, EEG data should be correctly preprocessed and prepared via Feature Engineering (FE). In fact, the choice of FE techniques can make the difference between an unusable ML model and a simple, effective model. In other words, it can be said that FE is crucial, especially when using complex, non-stationary data such as EEG. To this aim, in this paper we present a Systematic Mapping Study (SMS) focused on FE from EEG data used to identify mental disorders. Our SMS covers more than 900 papers, making it one of the most comprehensive to date, to the best of our knowledge. We gathered the mental disorder addressed, all the FE techniques used, and the Artificial Intelligence (AI) algorithm applied for classification from each paper. Our main contributions are: (i) we offer a starting point for new researchers on these topics, (ii) we extract the most used FE techniques to classify mental disorders, (iii) we show several graphical distributions of all used techniques, and (iv) we provide critical conclusions for detecting mental disorders. To provide a better overview of existing techniques, the FE process is divided into three parts: (i) signal transformation, (ii) feature extraction, and (iii) feature selection. Moreover, we classify and analyze the distribution of existing papers according to the mental disorder they treat, the FE processes used, and the ML techniques applied. As a result, we provide a valuable reference for the scientific community to identify which techniques have been proven and tested and where the gaps are located in the current state of the art.

Context. Automated analyses of Gaia astrometric data have led to the discovery of many new high-quality open cluster candidates. When a good determination of their parameters is available, these objects become excellent tools for investigating the properties of our Galaxy. Aims. We explore whether young open clusters can be readily identified from Gaia data alone by studying the properties of their Gaia colour-magnitude diagrams. We also want to compare the results of a traditional cluster analysis with those of automated methods. Methods. We selected three young open cluster candidates from the UBC catalogue, ranging from a well-populated object with a well-defined sequence to a poorly populated and poorly defined candidate. We obtained classification spectra for the brightest stars in each. We redetermined members based on EDR3 data and fitted isochrones to derive the age, distance, and reddening. Results. All three candidates are real clusters with ages below 100 Ma. UBC 103 is a moderately populous cluster, with an age around 70 Ma. At a distance of ∼3 kpc, it forms a binary cluster with nearby NGC 6683. UBC 114 is a relatively proximal (∼1.5 kpc) poorly populated cluster containing two early-B stars. UBC 587 is a dispersed and very young (≤10 Ma) cluster located at ∼3 kpc, behind the Cygnus X region, which may be a valuable tracer of the Orion arm. Conclusions. The OCfinder methodology for the identification of new open clusters is extremely effective, with even poor candidates resulting in interesting detections. The presence of an almost vertical photometric sequence in the Gaia colour-magnitude diagram is a safe way to identify young open clusters. Automated methods for the determination of cluster properties give approximate solutions, but these are still subject to difficulties. There is some evidence suggesting that artificial intelligence systems may systematically underestimate extinction, which may impact in the age determination.

Few-shot techniques offer a promising avenue to reduce the high demand for annotated data required by current machine learning-based applications, such as Optical Music Recognition (OMR). This is a field dedicated to the automatic transcription of music notation from sheet music images. Traditional OMR systems strongly depend on layout analysis, a crucial step involving the identification and segmentation of several components within a music score, such as staff lines, text, or notes. The standard approach requires extensive fully annotated training data, which are resource-intensive and time-consuming to label and curate by domain experts. We present a preliminary study on the use of few-shot learning to alleviate the disadvantages associated with manual annotations. The proposal minimizes the human effort required by employing only partial annotations. For this, we introduce an oversampling technique to train models using a limited set of annotated patches extracted from the score images. Our experimental findings, conducted on four benchmark datasets, underscore the efficacy of the proposed patch extraction. Despite operating with a reduced amount of annotated data, our method achieves performance levels competitive with models trained on the complete dataset. This work points out the potential of few-shot learning in the context of layout analysis for music scores, offering the promise of more efficient and accessible OMR systems.

Sentinel−1 (S1) data enables effective monitoring of displacements using persistent scatterer interferometry (PSI). S1 includes VV and VH polarization channels, allowing us to apply polarimetric techniques to PSI. In short, polarimetric PSI (PolPSI) exploits the available polarization channels to enhance the identification and processing of measurement points including persistent scatterers (PS) and distributed scatterers (DS). Previous works have shown the benefits of using PolPSI for PS points with S1 data, but the corresponding analysis for DS is missing. DS points are processed by finding a neighborhood of statistically homogeneous pixels (SHP) and averaging the phase within that neighborhood. In this work we show how dual-polarimetric data are stricter on the selection of the SHP group than single-polarimetric data. Thanks to the information added by the second channel, different land covers are not mixed in the SHP group. As a result, the number of points in the SHP groups is generally smaller than with VV alone, but they are more reliable. The impact of this strategy on the resulting deformation estimates is also investigated in this work, showing that the deformation areas are fully preserved and the influence of nearby pixels associated with other scene elements is avoided.

Landslides are destructive geohazards to people and infrastructure, resulting in hundreds of deaths and billions of dollars of damage every year. Therefore, mapping the rate of deformation of such geohazards and understanding their mechanics is of paramount importance to mitigate the resulting impacts and properly manage the associated risks. In this paper, the main outcomes relevant to the joint European Space Agency (ESA) and the Chinese Ministry of Science and Technology (MOST) Dragon-5 initiative cooperation project ID 59,339 “Earth observation for seismic hazard assessment and landslide early warning system” are reported. The primary goals of the project are to further develop advanced SAR/InSAR and optical techniques to investigate seismic hazards and risks, detect potential landslides in wide regions, and demonstrate EO-based landslide early warning system over selected landslides. This work only focuses on the landslide hazard content of the project, and thus, in order to achieve these objectives, the following tasks were developed up to now: a) a procedure for phase unwrapping errors and tropospheric delay correction; b) an improvement of a cross-platform SAR offset tracking method for the retrieval of long-term ground displacements; c) the application of polarimetric SAR interferometry (PolInSAR) to increase the number and quality of monitoring points in landslide-prone areas; d) the semiautomatic mapping and preliminary classification of active displacement areas on wide regions; e) the modeling and identification of landslides in order to identify triggering factors or predict future displacements; and f) the application of an InSAR-based landslide early warning system on a selected site. The achieved results, which mainly focus on specific sensitive regions, provide essential assets for planning present and future scientific activities devoted to identifying, mapping, characterizing, monitoring and predicting landslides, as well as for the implementatio

Forest canopy height estimation is essential for forest management and biomass estimation. In this study, we aimed to evaluate the capacity of TanDEM-X interferometric synthetic aperture radar (InSAR) data to estimate canopy height with the assistance of an external digital terrain model (DTM). A ground-to-volume ratio estimation model was proposed so that the canopy height could be precisely estimated from the random-volume-over-ground (RVoG) model. We also refined the RVoG inversion process with the relationship between the estimated penetration depth (PD) and the phase center height (PCH). The proposed method was tested by TanDEM-X InSAR data acquired over relatively homogenous coniferous forests (Teruel test site) and coniferous as well as broadleaved forests (La Rioja test site) in Spain. Comparing the TanDEM-X-derived height with the LiDAR-derived height at plots of size 50 m × 50 m, the root-mean-square error (RMSE) was 1.71 m (R2 = 0.88) in coniferous forests of Teruel and 1.97 m (R2 = 0.90) in La Rioja. To demonstrate the advantage of the proposed method, existing methods based on ignoring ground scattering contribution, fixing extinction, and assisting with simulated spaceborne LiDAR data were compared. The impacts of penetration and terrain slope on the RVoG inversion were also evaluated. The results show that when a DTM is available, the proposed method has the optimal performance on forest height estimation.

Nowadays, there are many different industrial processes in which people spend several hours performing tedious and repetitive tasks. Furthermore, most of these processes involve the manipulation of dangerous materials or machinery, such as the toy manufacturing, where people handle ovens with high temperatures and make weary physical effort for a long period of time during the process. In this work, it is presented an automatic and innovative collaborative robotic system that is able to deal with the demoulding task during the manufacturing process of toy dolls. The intelligent robotic system is composed by an UR10e robot with a RealSense RGB-D camera integrated which detects the pieces in the mould using a developed vision-based algorithm and extracts them by means of a custom gripper located and the end of the robot. We introduce a pipeline to perform the demoulding task of different plastic pieces relying in the use of this intelligent robotic system. Finally, to validate this approach, the automatic method has been successfully implemented in a real toy factory providing a novel approach in this traditional manufacturing process. The paper describes the robotic system performance using different forces and velocities, obtaining a success rate of more than 90% in the experimental results.