Rendering 3D virtual scenarios has become a popular alternative for generating per-pixel-labeled image datasets, especially in fields like autonomous driving. The approach is valuable for training neural perception models, such as semantic segmentation models, particularly when data might be scarce, expensive, or difficult to collect. However, fundamental questions persist within the research community regarding the generation and processing of these synthetic images, particularly a better understanding of the key factors influencing the performance of deep learning models trained with such synthetic images. In response, we conducted a series of experiments to elucidate the impact that common aspects involved in the generation of rendered synthetic images may have on the performance of neural semantic segmentation tasks. Our study used a recent autonomous driving synthetic dataset as our main testbed, allowing us to investigate the effect of different approaches when modeling their geometric, material, and lighting details. We also studied the impact of rendering noise, typically produced by path-tracing algorithms, as well as the impact of using different color transformations and tonemapping algorithms.

We introduce UrbanSyn, a photorealistic dataset acquired through semi-procedurally generated synthetic urban driving scenarios. Developed using high-quality geometry and materials, UrbanSyn provides pixel-level ground truth, including depth, semantic segmentation, and instance segmentation with object bounding boxes and occlusion degree. It complements GTAV and Synscapes datasets to form what we coin as the 'Three Musketeers'. We demonstrate the value of the Three Musketeers in unsupervised domain adaptation for image semantic segmentation. Results on real-world datasets, Cityscapes, Mapillary Vistas, and BDD100K, establish new benchmarks, largely attributed to UrbanSyn. We make UrbanSyn openly and freely accessible (www.urbansyn.org).

En las últimas décadas se ha producido un cambio radical en el concepto de museo hacia la valorización de la experiencia como vía de adquisición de conocimiento. Este cambio ha generado un entorno propicio para la entrada de instalaciones digitales interactivas que aúnan información y educación con una perspectiva lúdica que, con frecuencia, le lleva a trasladar elementos y características de los videojuegos para una experiencia más atractiva, inmersiva y memorable. El objeto del presente artículo ha sido redefinir elementos del diseño de videojuegos y su intersección con el diseño centrado en el usuario para su aplicación en el contexto museístico.

This paper introduces a novel procedural modeling system for generating 3D highway models, leveraging real-world data inputs such as highway layouts and manual annotations of essential elements. Our highly parameterized system facilitates easy customization, including modifications to the number of lanes and other key features, broadening its applicability across diverse domains, from urban planning and transportation engineering to virtual simulations.

Deep neural networks are well known for demanding large amounts of training data, motivating the appearance of multiple synthetic datasets covering multiple domains. However, synthetic datasets have not yet outperformed real data for autonomous driving applications, particularly for semantic segmentation tasks. Thus, a deeper comprehension about how the parameters involved in synthetic data generation could help in creating better synthetic datasets. This work provides a summary review of prior research covering how image noise, camera noise and rendering photorealism could affect learning tasks. Furthermore, we presents novel experiments aimed at advancing our understanding around generating synthetic data for autonomous driving neural networks aimed at semantic segmentation

The aim of the paper is to determine the elements, components and key facts to design interactive interfaces for augmented reality environments, focusing on the design and development of virtual worlds for educational applications. The article focuses on the usability study of three augmented reality applications interfaces developed for non touchable natural environments. The aim is to determine the menu and interactive items the user accesses without difficulty and also their navigation patterns. The most viewed elements and the ones that can't be seen by the user, generating problems of navigation and usability, are analyzed too. The research results indicate that applications developed for non-touch environments can be stressful for the user, especially because of the time they spend with their hands up. The space around the student is also important for the learning process, and all the applications must be designed with some educational, rather than leisure tasks, or a fully open navigation.

The aim of the paper is to determine the elements, components and key facts to design interactive interfaces for augmented reality environments, focusing on the design and development of virtual worlds for educational applications.The article focuses on the usability study of three augmented reality applications interfaces developed for non touchable natural environments. The aim is to determine the menu and interactive items the user accesses without difficulty and also their navigation patterns. The most viewed elements and the ones that can´t be seen by the user, generating problems of navigation and usability, are analyzed too. The research results indicate that applications developed for non-touch environments can be stressful for the user, especially because of the time they spend with their hands up. The space around the student is also important for the learning process, and all the applications must be designed with some educational, rather than leisure tasks, or a fully open navigation.

Este estudio analiza la visibilidad que tiene la publicidad insertada a modo de product placement en los videojuegos. Para ello se ha realizado una investigación cualitativa experimental mediante la tecnología de eye tracking que determinó las zonas de pantalla más visibles y eficientes para la inserción de marcas, en un entorno eminentemente lúdico. Esta investigación se centra en la eficacia y el recuerdo de las ubicaciones en nueve zonas de pantalla correspondientes a tres áreas verticales (derecha, central e izquierda) y tres áreas horizontales (superior, inferior y central). El objetivo es averiguar cuáles son las de mayor visibilidad y su asociación con el recuerdo, por parte del usuario, de cada una de las marcas anunciantes.

Draping 2D vectorial information over a 3D terrain elevation model is usually performed by real-time rendering to texture. In the case of linear feature representation, there are several specific problems using the texturing approach, specially when using multi-resolution textures. These problems are related to visual quality, aliasing artifacts and rendering performance. In this paper, we address the problems of 2D line rasterization on a multi-resolution texturing engine from a pragmatical point of view; some alternative solutions are presented, compared and evaluated. For each solution we have analyzed the visual quality, the impact on the rendering performance and the memory consumption. The study performed in this work is based on an OpenGL implementation of a clipmap-based multi-resolution texturing system, and is oriented towards the use of inexpensive consumer graphics hardware.

Immersive virtual reality constitutes a powerful tool for the vivid exploration of virtual spaces. The feeling of presence produced in the user by the action of seeing the digital space just pointing the view in any direction is greatly enhanced when adding the capability of walking physically during the experience. This paper describes aspects related with the experimentation of this hybrid space, real and virtual at the same time, that was implemented by the authors in an immersive virtual reality museum installation called "The Empty Museum".

Geographic Information Systems (GIS) provide large information management capabilities. However, in many cases the feature of the visualization of this data over a huge detailed three-dimensional terrain does not exist or are limited, wich can be very useful or desirable in many applications. This paper presents the integration of SANTI, a three-dimensional visualization system of large terrain areas and geographical data, with gvSIG, a free GIS software. We describe the capabilities of the visualization system, the approach to connect the two systems and how it is applied to a project in development for the management of the excavation of the Castro de la Lanzada located at the Galician coast (Spain).

Geographic Information Systems (GIS) provide large information management capabilities. However, in many cases the feature of the visualization of this data over a huge detailed three-dimensional terrain does not exist or are limited, wich can be very useful or desirable in many applications. This paper presents the integration of SANTI, a three-dimensional visualization system of large terrain areas and geographical data, with gvSIG, a free GIS software. We describe the capabilities of the visualization system, the approach to connect the two systems and how it is applied to a project in development for the management of the excavation of the Castro de la Lanzada located at the Galician coast (Spain).

Immersive virtual reality constitutes a powerful tool for the vivid exploration of virtual spaces. The feeling of presence produced in the user by the action of seeing the digital space just pointing the view in any direction is greatly enhanced when adding the capability of walking physically during the experience. This paper describes aspects related with the experimentation of this hybrid space, real and virtual at the same time, that was implemented by the authors in an immersive virtual reality museum installation called "The Empty Museum".

The real-time rendering of arbitrarily large textures is a problem that has long been studied in terrain visualization. For years, different approaches have been published that have either expensive hardware requirements or other severe limitations in quality, performance or versatility. The biggest problem is usually a strong coupling between geometry and texture, both regarding database structure, as well as LOD management. This paper presents a new approach to high resolution, real-time texturing of dynamic data that avoids the drawbacks of previous techniques and offers additional possibilities. The most important benefits are: out-of-core texture visualization from dynamic data, efficient per-fragment texture LOD computation, total independence from the geometry engine, high quality filtering and easiness of integration with user custom shaders and multitexturing. Because of its versatility and independence from geometry, the proposed technique can be easily and efficiently applied to any existing terrain geometry engine in a transparent way.

Graphics Processing Units (GPUs) have been evolving very fast, turning into high performance programmable processors. Though GPUs have been designed to compute graphics algorithms, their power and flexibility makes them a very attractive platform for generalpurpose computing. In the last years they have been used to accelerate calculations in physics, computer vision, artificial intelligence, database operations, etc. (Owens, 2007). In this paper an approach to general purpose computing with GPUs is made, followed by a description of artificial intelligence algorithms based on Artificial Neural Networks (ANN) and Evolutionary Computation (EC) accelerated using GPU.

RESUMEN En este artículo se describen las experiencias de los autores en el diseño y la ejecución de tres instalaciones interactivas para museos basados en el concepto de interfaces naturales, es decir, aquellos que hacen uso de los medios de comunicación utilizados por los seres humanos en su relación natural con su entorno a través de capacidades comunes tales como hablar, gesticular, caminar o tocar. Estas instalaciones son parte de la exposición permanente Galicia Dixital, en Santiago de Compostela, que se dedica a ilustrar sobre la cultura de esta región española al tiempo que introduce al visitante en las aplicaciones de las nuevas tecnologías. Palabras clave: Interfaces naturales, contenidos digitales interactivos, museos. ABSTRACT This paper describes the authors' experience in the design and implementation of three interactive installations for museums based in the natural interfaces concept; that is, those that make use of the ways of communication used by humans in their natural relation with their environment by means of common abilities such as talking, gesturing, walking or touching. These installations are part of the Galicia Dixital permanent exhibition in Santiago de Compostela, which is devoted to illustrate on the culture of this Spanish region while introducing the visitor in the applications of new technologies.

Immersive Virtual Reality Systems have been extensively used during recent years for the exploration of architectonic spaces. This paper describes how the use of transitable immersive virtual reality systems, that is, those that allow the user to physically walk while exploring the virtual world, can greatly empower the experience of perception of space in architecture. The text describes a particular example of one installation of this kind that was developed by the authors and how it was implemented for the interactive experience of the virtual reconstruction of a housing unit on a preroman settlement. This installation is open to the public as part of a permanent exhibition and constitutes the final output of the research at this time.

We present a technique for efficient management of large textures and its real-time application to geometric models. The proposed technique is inspired by the clipmap [12] idea, that caches in video memory a subset of the texture mipmap pyramid. Based on this concept, we define some structures and a different management allowing its implementation on a personal computer without specific graphics hardware. Finally, we present the results of the application in a terrain visualization system, using several simultaneous textures with a detail up to 0.25 meters per texel, covering a 60,000 km(2) area.