We present the development of an automatic dental diagnosis system based on artificial intelligence. The goal is to accurately support clinical decision-making in dentistry based on pantomographic images. We describe the construction of a large dataset of pantomographic images with high-quality annotations, the development of an intelligent model for semantic image interpretation and generation of presumptive diagnoses, and an application for end users. The model demonstrates the ability to correctly identify each individual tooth, detect multiple conditions and common pathologies in dental pieces, in case they arise, and generate a presumptive diagnostic text based on the information extracted from the image. The application, in the form of a digital web-based negatoscope, represents a technological advancement in radiology in line with the digital transformation era, offering a versatile and efficient tool for the interactive visualization of radiographic images, especially designed for dental professionals.

Autonomous driving, decades ago relegated to the realm of science fiction, emerged as a tangible reality that is rapidly transforming the automotive industry, redefining our relationship with vehicles, and placing them in the spotlight of both the industry and the general public. Through the study and analysis of modern and efficient interpolation techniques, we aim to reduce the current costs and processing requirements associated with the LiDAR sensor, which is one of the main information sources. Our approach explores the fusion of lower-cost LiDAR sensors with advanced interpolation techniques, with a particular focus on achieving performance parity with pricier 64-channel LiDAR setups. This work is based on 3 main axes: firstly, the analysis of available LiDAR data and its representation; secondly, the development and implementation of an interpolation technique based on 1D convolutional layers integrated with fully connected layers, in order to analyse data coming from a sliding window; and finally, the comparative evaluation of the results between different state-of-the-art interpolation techniques, using object detection networks in point clouds. By interpolating the point clouds with the proposed technique, improvements between 1.92% and 30.98% in detection and classification tasks were achieved, depending on the object and the type of detection (3D or bird's eye view). Furthermore, computational efficiency was not left aside by reducing the inference times necessary for interpolation, compared to other techniques used as contrast. This highlights the viability and scalability of our approach in realizing cost-effective yet high-perfermance autonomous driving systems.

The popularization of unmanned aerial vehicles (UAVs) is transforming contemporary Geotechnology related activities, offering accessibility, accuracy, and efficiency. In particular, landcover analysis and cartography with UAV-borne imagery enables precise identification and measurements for surveying purposes. While multispectral cameras facilitate detailed mapping, affordability drives the use of common RGB drone cameras. This, in turn, raises the requirement of robust and versatile analysis techniques for extracting meaningful information from RGB images. In this study, we assess different algorithms for mapping urban areas using high-resolution RGB aerial images from UAVs. We explore the use of different machine learning methods in landcover identification and classification, leading to a workflow for producing thematic maps, showcasing the potential of machine learning in urban mapping with UAV-acquired RGB imagery.

Accurate completion of archaeological artifacts is a critical aspect in several archaeological studies, including documentation of variations in style, inference of chronological and ethnic groups, and trading routes trends, among many others. However, most available pottery is fragmented, leading to missing textural and morphological cues. Currently, the reassembly and completion of fragmented ceramics is a daunting and time-consuming task, done almost exclusively by hand, which requires the physical manipulation of the fragments. To overcome the challenges of manual reconstruction, reduce the materials' exposure and deterioration, and improve the quality of reconstructed samples, we present IberianVoxel, a novel 3D Autoencoder Generative Adversarial Network (3D AE-GAN) framework tested on an extensive database with complete and fragmented references. We generated a collection of 1001 3D voxelized samples and their fragmented references from Iberian wheel-made pottery profiles. The fragments generated are stratified into different size groups and across multiple pottery classes. Lastly, we provide quantitative and qualitative assessments to measure the quality of the reconstructed voxelized samples by our proposed method and archaeologists' evaluation.

Ergonomic risk assessment is now, due to an increased awareness, carried out more often than in the past. The conventional risk assessment evaluation, based on expert-assisted observation of the workplaces and manually filling in score tables, is still predominant. Data analysis is usually done with a focus on critical moments, although without the support of contextual information and changes over time. In this paper we introduce ErgoExplorer, a system for the interactive visual analysis of risk assessment data. In contrast to the current practice, we focus on data that span across multiple actions and multiple workers while keeping all contextual information. Data is automatically extracted from video streams. Based on carefully investigated analysis tasks, we introduce new views and their corresponding interactions. These views also incorporate domain-specific score tables to guarantee an easy adoption by domain experts. All views are integrated into ErgoExplorer, which relies on coordinated multiple views to facilitate analysis through interaction. ErgoExplorer makes it possible for the first time to examine complex relationships between risk assessments of individual body parts over long sessions that span multiple operations. The newly introduced approach supports analysis and exploration at several levels of detail, ranging from a general overview, down to inspecting individual frames in the video stream, if necessary. We illustrate the usefulness of the newly proposed approach applying it to several datasets.

Abstract Finding point‐wise correspondences between images is a long‐standing problem in image analysis. This becomes particularly challenging for sketch images, due to the varying nature of human drawing style, projection distortions and viewport changes. In this paper, we present the first attempt to obtain a learned descriptor for dense registration in line drawings. Based on recent deep learning techniques for corresponding photographs, we designed descriptors to locally match image pairs where the object of interest belongs to the same semantic category, yet still differ drastically in shape, form, and projection angle. To this end, we have specifically crafted a data set of synthetic sketches using non‐photorealistic rendering over a large collection of part‐based registered 3D models. After training, a neural network generates descriptors for every pixel in an input image, which are shown togeneralize correctly in unseen sketches hand‐drawn by humans. We evaluate our method against a baseline of correspondences data collected from expert designers, in addition to comparisons with other descriptors that have been proven effective in sketches. Code, data and further resources will be publicly released by the time of publication.

Abstract Following Archimedes' Principle, any object immersed in a fluid is subject to an upward buoyancy force equal to the weight of the fluid displaced by the object. This simple description is the origin of a set of effects that are ubiquitous in nature, and are becoming commonplace in games, simulators and interactive animations. Although there are solutions to the fluid‐to‐solid coupling problem in some particular cases, to the best of our knowledge, comprehensive and accurate computational buoyancy models adequate in general contexts are still lacking. We propose a real‐time Graphics Processing Unit (GPU) based algorithm for realistic computation of the fluid‐to‐solid coupling problem, which is adequate for a wide generality of cases (solid or hollow objects, with permeable or leak‐proof surfaces, and with variable masses). The method incorporates the behaviour of the fluid into which the object is immersed, and decouples the computation of the physical parameters involved in the buoyancy force of the empty object from the mass of contained liquid. The dynamics of this mass of liquid are also computed, in a way such that the relation between the centre of mass of the object and the buoyancy force may vary, leading to complex, realistic beha viours such as the ones arising for instance with a sinking boat.

The development of data approximation methods from coarse spatial resolution images is gaining increasing interest in the research community. This letter aims to extend and validate a developed methodology for estimating border length of diverse marine coastlines from coarse spatial resolution images like Moderate Resolution Imaging Spectrometer (MODIS) by using fractal attributes and its error behavior. The accuracy of MODIS-based estimates and the reliability of the method to predict coastline length measurements by extrapolation was evaluated using Landsat 8 over different coastline types. It is shown that with our method, 250-m MODIS images are adequate for estimating coastline lengths with a precision equivalent to standard linear measurements performed on 30-m resolution imagery, with average errors between 3% and 18% for straight and complex coasts, respectively. These results indicate that an underestimation error, occurring in rugged and complex coasts, is more frequent and significant than overestimation occurring in smooth and straight coasts.

Radial charts are generally considered less effective than linear charts. Perhaps the only exception is in visualizing periodical time-dependent data, which is believed to be naturally supported by the radial layout. It has been demonstrated that the drawbacks of radial charts outweigh the benefits of this natural mapping. Visualization of daily patterns, as a special case, has not been systematically evaluated using radial charts. In contrast to yearly or weekly recurrent trends, the analysis of daily patterns on a radial chart may benefit from our trained skill on reading radial clocks that are ubiquitous in our culture. In a crowd-sourced experiment with 92 non-expert users, we evaluated the accuracy, efficiency, and subjective ratings of radial and linear charts for visualizing daily traffic accident patterns. We systematically compared juxtaposed 12-hours variants and single 24-hours variants for both layouts in four low-level tasks and one high-level interpretation task. Our results show that over all tasks, the most elementary 24-hours linear bar chart is most accurate and efficient and is also preferred by the users. This provides strong evidence for the use of linear layouts - even for visualizing periodical daily patterns.

The authors propose a video‐summarisation method based on visual and categorical diversities using pre‐trained deep visual and categorical models. Their method extracts visual and categorical features from a pre‐trained deep convolutional network (DCN) and a pre‐trained word‐embedding matrix. Using visual and categorical information they obtain a video diversity estimation, which is used as an importance score to select segments from the input video that best describes it. Their method also allows performing queries during the search process, in this way personalising the resulting video summaries according to the particular intended purposes. The performance of the method is evaluated using different pre‐trained DCN models in order to select the architecture with the best throughput. They then compare it with other state‐of‐the‐art proposals in video summarisation using a data‐driven approach with the public dataset SumMe, which contains annotated videos with per‐fragment importance. The results show that their method outperforms other proposals in most of the examples. As an additional advantage, their method requires a simple and direct implementation that does not require a training stage.

In neuroimaging, magnetic resonance images can be used to locate and obtain various parameters in order to find a wide range of pathologies, improving diagnosis and hence early treatment. Since images of the brain are volumetric, they are treated volumetrically in voxels, rather than planarly in pixels. We present an alternative implementation of local neighborhood-based texture parameters that have been recently shown to improve the detection of differences in the brain between healthy patients and those with Alzheimer's disease using diffusion tensor imaging [1]. We implemented the method (1) in Python using the Anaconda environment and the PyCharm compiler and (2) in 3D Slicer environment, as it is widely used by the neurology community, like the National Alliance for Medical Image Computing, among others. Weighted rotational invariant local operators were used in the calculus, namely average, standard deviation, coefficient of variation, normalized skewness, median, inter-quartile range and quartile coefficient of variation. Comparison between the implementations has been measured with normalized root-mean-square error. No differences have been observed for the non-linear parameters based on quartiles and errors smaller than 0.5% have been observed for operators that used Fast Fourier Transform based convolution instead of the explicit method.

Abstract Probabilistic weather forecasts are amongst the most popular ways to quantify numerical forecast uncertainties. The analog regression method can quantify uncertainties and express them as probabilities. The method comprises the analysis of errors from a large database of past forecasts generated with a specific numerical model and observational data. Current visualization tools based on this method are essentially automated and provide limited analysis capabilities. In this paper, we propose a novel approach that breaks down the automatic process using the experience and knowledge of the users and creates a new interactive visual workflow. Our approach allows forecasters to study probabilistic forecasts, their inner analogs and observations, their associated spatial errors, and additional statistical information by means of coordinated and linked views. We designed the presented solution following a participatory methodology together with domain experts. Several meteorologists with different backgrounds validated the approach. Two case studies illustrate the capabilities of our solution. It successfully facilitates the analysis of uncertainty and systematic model biases for improved decision‐making and process‐quality measurements.

Form and shape descriptors are among the most useful features for object identification and recognition. Even though there exists a widely used set of shape descriptors and underlying computational methods for their evaluation, frequently they fail to distinguish among very similar objects that they appear very different to the human eye. The authors propose a new set of shape descriptors based on a finer determination of the object symmetry axes, and a more accurate estimation of the bilateral and radial symmetries. These descriptors were thoroughly tested using several synthetic and real objects with varying degrees of symmetry. The methods for axes estimation and symmetry descriptors extraction outperform the widespread shape descriptors in recognising and identifying among very similar objects.

Abstract Photorealistic modeling and rendering of materials with complex internal mesostructure is a hard challenge in Computer Graphics. In particular, macroscopic porous materials consist of complex translucent substances that exhibit different details and light interaction at several different scales. State‐of‐the‐art techniques for modeling porous materials manage the material either as a surface and set up complex capture procedures or as a volume by employing different instances of procedural noise models for its representation. While the surface solution achieves several desired material properties, it still presents drawbacks in practical applications—high computational costs, complex capture procedures, and poor image variability, among others. Volumetric solutions are more flexible, but the final structure and appearance are difficult to control. To overcome these drawbacks, we propose an algorithm for the procedural generation of porous materials. The method is based on an artistic and physically inspired simulation of the growth of self‐avoiding bubbles inside a volume, by means of dynamical systems. The patterns induced by the bubbles can be easily and intuitively controlled. The bubbles adapt to any given shape and have convincing global and local fluid‐like patterns as seen in bread and sponges. Our method generates 3D textures that adequately represent porous materials, which can be used as input for creating realistic renderings of different porous objects. As a case study, we present the results of using these 3D textures as input to a direct volume renderer and show that they compare favorably with standard 3D texture synthesis methods. Copyright © 2016 John Wiley & Sons, Ltd.

Public transportation movement data provide a wealth of information and insights into many aspects of urban life and human behavior. However, huge amounts of raw data, coupled with incomplete or inconsistent records, may turn into an obstacle for the effective use of the available information. The need for effective movement data analysis has resulted in a large number of visual analytics tools and specialized views. There are still many challenges in public transportation and other kinds of cyclic movement data analysis. In this paper we address some of those challenges by presenting an improvement of the standard map view. This improved view is specifically designed to simplify and make the visual analysis of complex movement data easier to perform, especially when integrated in a coordinated multiple views tool and articulated together with other techniques. We illustrate the effectiveness of the view on public transportation data from Bahía Blanca, Argentina.

Abstract Weather conditions affect multiple aspects of human life such as economy, safety, security, and social activities. For this reason, weather forecast plays a major role in society. Currently weather forecasts are based on Numerical Weather Prediction (NWP) models that generate a representation of the atmospheric flow. Interactive visualization of geo‐spatial data has been widely used in order to facilitate the analysis of NWP models. This paper presents a visualization system for the analysis of spatio‐temporal patterns in short‐term weather forecasts. For this purpose, we provide an interactive visualization interface that guides users from simple visual overviews to more advanced visualization techniques. Our solution presents multiple views that include a timeline with geo‐referenced maps, an integrated webmap view, a forecast operation tool, a curve‐pattern selector, spatial filters, and a linked meteogram. Two key contributions of this work are the timeline with geo‐referenced maps and the curve‐pattern selector. The latter provides novel functionality that allows users to specify and search for meaningful patterns in the data. The visual interface of our solution allows users to detect both possible weather trends and errors in the weather forecast model. We illustrate the usage of our solution with a series of case studies that were designed and validated in collaboration with domain experts.

Digital video has become a very popular media in several contexts, with an ever expanding horizon of applications and uses. Thus, the amount of available video data is growing almost limitless. For this reason, video summarization continues to attract the attention of a wide spectrum of research efforts. In this work we present a novel video summarization technique based on tracking local features among consecutive frames. Our approach operates on the uncompressed domain, and requires only a small set of consecutive frames to perform, thus being able to process the video stream directly and produce results on the fly. We tested our implementation on standard available datasets, and compared the results with the most recent published work in the field. The results achieved show that our proposal produces summarizations that have similar quality than the best published proposals, with the additional advantage of being able to process the stream directly in the uncompressed domain.