Realistic avatars are fundamental for immersive experiences in Augmented Reality (AR) and Virtual Reality (VR) environments. In this work, we introduce a novel approach for avatar generation, combining 3D Gaussian Spl...
详细信息
ISBN:
(纸本)9798350374490;9798350374506
Realistic avatars are fundamental for immersive experiences in Augmented Reality (AR) and Virtual Reality (VR) environments. In this work, we introduce a novel approach for avatar generation, combining 3D Gaussian Splatting with the parametric body model, SMPL. This methodology overcomes the inefficiencies of traditional image/video-based avatar creation, which is often slow and requires high computing resources. The integration of 3D Gaussian Splatting for representing human avatar offers realistic and real-time rendering for AR/VR applications. We also conducted preliminary tests to verify the quality of avatar representation using 3D Gaussian Splatting. These tests, displayed alongside outcomes from existing methods, demonstrate the potential of this research to significantly contribute to the creation of realistic avatars in the future. Additionally, several key discussions are presented, essential for developing and evaluating the system and providing valuable insights for future research.
Traditionally, topics such as parallel computing, computer graphics, and artificial intelligence have been taught as stand-alone courses in the computing curriculum. Often these are elective courses, limiting the mate...
详细信息
ISBN:
(纸本)9781479931903
Traditionally, topics such as parallel computing, computer graphics, and artificial intelligence have been taught as stand-alone courses in the computing curriculum. Often these are elective courses, limiting the material to the subset of students choosing to take the course. Recently there has been movement to distribute topics across the curriculum in order to ensure that all graduates have been exposed to concepts such as parallel computing. This work describes one approach to threading topics from AI, graphics, and parallel processing into course-work starting early in a student's program without substantially altering the basic course objectives. This, in turn, allows the student to sample topics early enough to make informed decisions when selecting electives.
Physically-Based Rendering (PBR) is key to modeling the interaction between light and materials, and finds extensive applications across computer graphics domains. However, acquiring PBR materials is costly and requir...
详细信息
ISBN:
(纸本)9798350353013;9798350353006
Physically-Based Rendering (PBR) is key to modeling the interaction between light and materials, and finds extensive applications across computer graphics domains. However, acquiring PBR materials is costly and requires special apparatus. In this paper, we propose a method to extract PBR materials from a single real-world image. We do so in two steps: first, we map regions of the image to material concept tokens using a diffusion model, allowing the sampling of texture images resembling each material in the scene. Second, we leverage a separate network to decompose the generated textures into spatially varying BRDFs (SVBRDFs), offering us readily usable materials for rendering applications. Our approach relies on existing synthetic material libraries with SVBRDF ground truth. It exploits a diffusion-generated RGB texture dataset to allow generalization to new samples using unsupervised domain adaptation (UDA). Our contributions are thoroughly evaluated on synthetic and real-world datasets. We further demonstrate the applicability of our method for editing 3D scenes with materials estimated from real photographs. Along with video, we share code and models as open-source on the project page: https://***/astra-vision/MaterialPalette
In computer graphics, 3D modeling is a fundamental concept. It is the process of creating three-dimensional objects or scenes using specialized software that allows users to create, manipulate and modify geometric sha...
详细信息
ISBN:
(纸本)9783031500718;9783031500725
In computer graphics, 3D modeling is a fundamental concept. It is the process of creating three-dimensional objects or scenes using specialized software that allows users to create, manipulate and modify geometric shapes to build complex models. This operation requires a huge amount of time to perform and specialised knowledge. Typically, it takes three to five hours of modelling to obtain a basic mesh from the blueprint. Several approaches have tried to automate this operation to reduce modelling time. The most interesting of these approaches are based on Deep Learning, and one of the most interesting is Pixel2Mesh. However, training this network requires at least 150 epochs to obtain usable results. Starting from these premises, this work investigates the possibility of training a modified version of the Pixel2Mesh in fewer epochs to obtain comparable or better results. A modification was applied to the convolutional block to achieve this, replacing the classification-based approach with an image reconstruction-based approach. This modification uses a configuration based on constructing an encoder-decoder architecture using state-of-the-art networks such as VGG, DenseNet, ResNet, and Inception. Using this approach, the convolutional block learns how to reconstruct the image correctly from the source image by learning the position of the object of interest within the image. With this approach, it was possible to train the complete network in 50 epochs, achieving results that outperform the state-of-the-art. The tests performed on the networks show an increase of 0.5% points over the state-of-the-art average.
Tooth preparation is a fundamental treatment technique to restore oral function in prosthodontic dentistry. This technique is complicated as it requires the preparation of an abutment while simultaneously predicting t...
详细信息
ISBN:
(纸本)9798350374490;9798350374506
Tooth preparation is a fundamental treatment technique to restore oral function in prosthodontic dentistry. This technique is complicated as it requires the preparation of an abutment while simultaneously predicting the ideal shape. We explore the feasibility of using Augmented Reality (AR) Head-Mounted Displays (HMDs) to assist dentists during tooth preparation using two different visualization techniques. A user study (N=24) revealed that AR is effective for angle adjustment, and reduces the occurrence of over-reduction. These results suggest that AR can be used to assist physicians during these procedures and has the potential to enhance the accuracy and safety of prosthodontic treatment.
The use of the primitive objects, can be modified and adapted for universal applications in computer graphics context. The present paper addresses some access and research in the field of computer graphics. The progra...
详细信息
ISBN:
(纸本)9789532330816
The use of the primitive objects, can be modified and adapted for universal applications in computer graphics context. The present paper addresses some access and research in the field of computer graphics. The program code present the basics step to increase the study and to explore differently level by describing each of them by math formulas and graphical representations. Also, here is explained the methods how we can use all these recourses to implement to any system, where can be find the better solution in the context of universal applications and solutions. So, to solve all from above questions, are given some characteristics objects and program code for respective representations. Also is given a direct approach for ADC (analog-digital conversion), by use the fundamental steps followed by this topics: sampling rate, quantization. Also is given statistical results by creating the respective images and respective view in the context of the statistical analyses.
This paper evaluates the digital surface model (DSM) generated from satellite stereo imageries that are simulated by the technology of 3-dimensional computer graphics (3D/CG), as one of the preliminary simulations for...
详细信息
ISBN:
(纸本)9798350360332;9798350360325
This paper evaluates the digital surface model (DSM) generated from satellite stereo imageries that are simulated by the technology of 3-dimensional computer graphics (3D/CG), as one of the preliminary simulations for the future mission in the Japan Aerospace Exploration Agency (JAXA). By using the 3D/CG instead of real satellite imageries, arbitrary conditions of stereo observation, such as the combinations of viewing angles, can be simulated with fewer limitations. Here we validate the effects of the B/H differences in the DSM generations as a first experiment of this approach. The height accuracies of the DSMs, which are derived from simulated stereo imageries of different B/Hs, are compared quantitatively. The result showed that height errors in the city area rapidly increase as the B/Hs are enlarged due to the effect of occlusions around the buildings.
Intrinsic decomposition of omnidirectional video is a challenging task. We propose a method that can provide temporally consistent decomposition results. Leveraging the 360-degree scene representation, we maintain the...
详细信息
ISBN:
(纸本)9798350374490;9798350374506
Intrinsic decomposition of omnidirectional video is a challenging task. We propose a method that can provide temporally consistent decomposition results. Leveraging the 360-degree scene representation, we maintain the global point cloud to propagate and reuse the similar inter-frame content and establish temporal constraints which elevate the quality of frame-wise decomposition while maintaining inter-frame coherence. By optimizing the proposed objective function, our method achieves a precise separation of reflectance and shading components. Comprehensive experiments demonstrate that our approach outperforms existing intrinsic decomposition methods. Our method also hold promise for various video manipulation applications.
Due to the lack of suitable phase distortion metrics, the optimization and evaluation of phase-holograms have relied on numerical reconstruction (NR) domain metrics. However, the necessity of NR for hologram processin...
详细信息
ISBN:
(纸本)9798350374490;9798350374506
Due to the lack of suitable phase distortion metrics, the optimization and evaluation of phase-holograms have relied on numerical reconstruction (NR) domain metrics. However, the necessity of NR for hologram processing leads to more intricate design and a higher computational burden. In this paper, we introduce a distribution-shifting (DS) algorithm to enable optimizing or measuring distortions directly in phase-domain, considering 2p-periodicity and shift-invariance. Experimental results with various noise types demonstrate a strong correlation between phase-domain metric with DS and their NR-domain counterparts. We believe that our DS-metric could facilitate direct optimization approaches in various phase-hologram processing techniques.
Sound propagation in 3D spaces is governed by similar physical principles as light. As a result, sound rendering in a 3D virtual environment can benefit from methods developed for graphics rendering and vice versa. In...
详细信息
ISBN:
(纸本)9781479972272
Sound propagation in 3D spaces is governed by similar physical principles as light. As a result, sound rendering in a 3D virtual environment can benefit from methods developed for graphics rendering and vice versa. In this review, we provide an overview of methods used for sound rendering that share concepts and techniques with graphics rendering. Firstly we describe geometrical propagation techniques where the computations are based on ray theory similar to ray tracing techniques in computer graphics. Secondly, we review numerical techniques. These techniques, similar to the idea of radiosity, are based on the subdivision of the space into elements. Then we describe acceleration techniques that can be used in combination with other methods to speed up calculations. Lastly, for the sake of completeness, a quick overview is given of sound computation techniques that simulate specific sound effects that do not apply on illumination. The aim of this survey is to share knowledge among the two disciplines using familiar and known concepts.
暂无评论