The pipeline for creating high-fidelity facial models often utilizes multi-view stereo techniques for reconstruction. However, the subsequent step of retopology often involves intricate manual work, limiting the exten...
详细信息
ISBN:
(纸本)9798350374490;9798350374506
The pipeline for creating high-fidelity facial models often utilizes multi-view stereo techniques for reconstruction. However, the subsequent step of retopology often involves intricate manual work, limiting the extension of facial capture systems towards 4D acquisition. This paper proposes a facial 4D capture pipeline based on high-speed cameras. We employ standard multi-view stereo techniques for 3D reconstruction. Non-linear deformations of facial expressions are decoupled from rigid movements of the skull using QR code markers. Additionally, a progressive automated retopology approach is introduced for batch processing. Results demonstrate that our system can capture continuous facial motion sequences with detailed 3D models.
Change detection(CD)is becoming indispensable for unmanned aerial vehicles(UAVs),especially in the domain of water landing,rescue and ***,even the most advanced models require large amounts of data for model training ...
详细信息
Change detection(CD)is becoming indispensable for unmanned aerial vehicles(UAVs),especially in the domain of water landing,rescue and ***,even the most advanced models require large amounts of data for model training and ***,sufficient labeled images with different imaging conditions are *** by computer graphics,we present a cloning method to simulate inland-water scene and collect an auto-labeled simulated *** simulated dataset consists of six challenges to test the effects of dynamic background,weather,and noise on change detection ***,we propose an image translation framework that translates simulated images to synthetic *** framework uses shared parameters(encoder and generator)and 22×22 receptive fields(discriminator)to generate realistic synthetic images as model training *** experimental results indicate that:1)different imaging challenges affect the performance of change detection models;2)compared with simulated images,synthetic images can effectively improve the accuracy of supervised models.
Recently, neural network-based approaches for hologram generation and compression have gained popularity as they allow for efficient inference on GPUs without the need for iterative optimization required in traditiona...
详细信息
ISBN:
(纸本)9798350374025;9798350374032
Recently, neural network-based approaches for hologram generation and compression have gained popularity as they allow for efficient inference on GPUs without the need for iterative optimization required in traditional methods. In this paper, we introduce Neural Holographic Video Compression (NHVC), an end-to-end trainable and scalable model designed for high-quality phase hologram video generation and compression. NHVC consists of an auto-encoder-based phase hologram generator, a latent coder and- two hyper-prior coders. For each input image, the latent features are extracted through the encoder part of the phase generator and then entropy coded at the shared latent coder based on the hyper-prior information. The two hyper-prior coders employ a spatial and a spatio-temporal entropy model for I-frames and P-frames, respectively. With this architecture, our NHVC can offer task-scalability, allowing a single trained model to serve as a phase hologram generator, phase hologram image compressor, or phase hologram video compressor as required. Experimental results on phase hologram video compression with UVG dataset show that our model outperforms 'HoloNet + VVC' by 75.6% BD-Rate reduction, with modest 2K encoding and decoding speeds (5 fps and 12 fps, respectively). For the phase hologram video generation task, our model showed much higher-quality (almost 42dB PSNR) reconstruction using the UVG dataset, while the previous neural generation model HoloNet provides at most 36dB reconstruction quality. We also provide an extensive experimental study on several important design questions such as the need for quadruple extension (QE) in the neural compression model, the feasibility of motion estimation in the phase domain, and an alternative, the need for increasing receptive field to learn better phase features, and variable rate support with a single trained model. It is noteworthy that our model is the first and best neural phase video compression model providing such high
We are witnessing a sudden increase in the general public's interest in computer graphics, whose creation is due to AI image generators. These tools are responsible for creating realistic graphics from the entered...
详细信息
ISBN:
(纸本)9783031662676;9783031662683
We are witnessing a sudden increase in the general public's interest in computer graphics, whose creation is due to AI image generators. These tools are responsible for creating realistic graphics from the entered text description. They are based on deep learning algorithms trained on an extensive graphical data set. Recently, the artificial intelligence system behind OpenAI, DALL-E 2, which can generate digital images based on textual descriptions, has become widely available. In addition, more and more similar tools are appearing on the market, such as Craiyon, Imagen, Midjourney, Stable Diffusion, etc. AI-generated designis an emerging technology whose advantages and disadvantages in engineering applications have yet to be verified. We analysed the advantages and disadvantages of applying this technology in product design within Industry 4.0.
Traditional 3D content creation tools empower users to bring their imagination to life by giving them direct control over a scene's geometry, appearance, motion, and camera path. Creating computer-generated videos...
详细信息
ISBN:
(纸本)9798350353013;9798350353006
Traditional 3D content creation tools empower users to bring their imagination to life by giving them direct control over a scene's geometry, appearance, motion, and camera path. Creating computer-generated videos, however, is a tedious manual process, which can be automated by emerging text-to-video diffusion models. Despite great promise, video diffusion models are difficult to control, hindering a user to apply their own creativity rather than amplifying it. To address this challenge, we present a novel approach that combines the controllability of dynamic 3D meshes with the expressivity and editability of emerging diffusion models. For this purpose, our approach takes an animated, low-fidelity rendered mesh as input and injects the ground truth correspondence information obtained from the dynamic mesh into various stages of a pre-trained text-to-image generation model to output high-quality and temporally consistent frames. We demonstrate our approach on various examples where motion can be obtained by animating rigged assets or changing the camera path. Project page: ***/generative_rendering.
In this study, we proposed a method to generate a more realistic image of buildings drawn by the three-point perspective method than the conventional perspective projection when the building is observed in real *** on...
详细信息
Locating neck-like features, or locally narrow parts, of a surface is crucial in various applications such as segmentation, shape analysis, path planning, and robotics. Topological methods are often utilized to find t...
详细信息
Locating neck-like features, or locally narrow parts, of a surface is crucial in various applications such as segmentation, shape analysis, path planning, and robotics. Topological methods are often utilized to find the set of shortest loops around handles and tunnels. However, there are abundant neck-like features on genus-0 shapes without any handles. While 3D geometry-aware topological approaches exist to find such loops, their construction can be cumbersome and may even lead to unintuitive loops. Here we present two methods for efficiently computing a complete set of surface loops that are not limited to the topologically nontrivial independent *** first approach is an efficient “topology-aware geometric” method to compute the tightest loops around neck features on surfaces, including genus-0 surfaces. We use the critical points of a processed distance function (such as Morse function) to find both the location and evaluate the significance of possible neck-like features. Critical points of a Morse function defined on a volume provide rich topological and geometric information about the structure of the shape. Our algorithm starts with a volumetric representation of an input surface and then calculates the distance function of mesh points to the boundary surface as a Morse function. We directly create a cutting plane through each neck feature. Each resulting loop can then be tightened to form a closed geodesic representation of the neck feature. Moreover, we offer criteria to measure the significance of a neck feature through the evolution of critical points during the smoothing of the distance function. Furthermore, we speed up the detection process through mesh simplification without compromising the quality of the output *** is known that reducing the dimension of a problem typically boosts efficiency drastically. Hence, we propose our second approach, which is a novel, efficient method that uses the skeleton of the shape to compute surface loops.
Physics-based simulations are a powerful tool in both computer graphics and engineering applications. Implicit discretization is essential for accurate, stable, and efficient simulations of solids and *** this thesis,...
详细信息
Physics-based simulations are a powerful tool in both computer graphics and engineering applications. Implicit discretization is essential for accurate, stable, and efficient simulations of solids and *** this thesis, we first present a novel implicit Material Point Method (MPM) discretization of spatially varying surface energies. Our discretization is based on surface energy, enabling implicit time stepping and capturing surface gradients without explicitly resolving them as in traction-condition-based approaches. We include an implicit discretization of thermomechanical material coupling with novel particle-based enforcement of Robin boundary conditions. Lastly, we design a particle resampling approach for perfect conservations of linear and angular momentum with Affine-Particle-In-Cell (APIC) [JSS15].The second part presents a novel deep-learning approach to approximate the solution of large, sparse, symmetric, positive-definite linear systems of equations. Our method is motivated by the conjugate gradients algorithm that iteratively selects search directions for minimizing the matrix norm of the approximation error. We use a deep neural network to accelerate convergence via data-driven improvement of the search direction at each iteration. We demonstrate the efficacy of our approach on discretized Poisson equations with millions of degrees of freedom. Our algorithm can reduce the linear system residual to the target tolerance in a small number of iterations, independent of the problem size, and generalize effectively to various systems beyond those encountered during ***, we present improvements to Position Based Dynamics (PBD) [MHH07] and Extended Position Based Dynamics (XPBD) [MMC16] methods, which are variants of implicit time integrator. PBD/XPBD are powerful methods for the real-time simulation of elastic objects, but they do not always converge. We isolate the root cause in the approximate linearization of the nonlinear backward Euler
In image morphing applications, deformation is not desired since it looks strange. To avoid deformation, the angles must be locally preserved. By deforming features, we can obtain funny-looking images, however, the sh...
详细信息
ISBN:
(纸本)9798350343557
In image morphing applications, deformation is not desired since it looks strange. To avoid deformation, the angles must be locally preserved. By deforming features, we can obtain funny-looking images, however, the shapes of the features will not look preserved if angles are not preserved. The angle-preserving operations are called conformal maps and locally they expect to be reduced into a concatenation of rotation, uniform scaling, and translation operations. This conceptual framework appears to give us a straightforward framework to construct general conformal maps. We can simply define desired rotation, uniform scaling, and translation in some select points and interpolate them by preserving their conformal property. Unfortunately, this does not work if these local operations are defined by using 3x3 matrices. Our approach is based on the well-known conformal property of maps in complex domains. Namely, any map in a complex domain is angle-preserving. Based on this well-known property of complex numbers, we show that if the local operations are defined as affine operations on complex numbers then any interpolation on complex numbers will give us conformal mapping of the whole domain. This is very useful since we can make extreme exaggeration of some local features and the resulting maps are still guaranteed to be conformal. Using this property, we have implemented an extended and conformal version of feature-based image metamorphosis.
computer-Generated Holography (CGH) algorithms simulate numerical diffraction, being applied in particular for holographic display technology. Due to the wave-based nature of diffraction, CGH is highly computationally...
详细信息
computer-Generated Holography (CGH) algorithms simulate numerical diffraction, being applied in particular for holographic display technology. Due to the wave-based nature of diffraction, CGH is highly computationally intensive, making it especially challenging for driving high-resolution displays in real-time. To this end, we propose a technique for efficiently calculating holograms of 3D line segments. We express the solutions analytically and devise an efficiently computable approximation suitable for massively parallel computing architectures. The algorithms are implemented on a GPU (with CUDA), and we obtain a 70-fold speedup over the reference point-wise algorithm with almost imperceptible quality loss. We report real-time frame rates for CGH of complex 3D line-drawn objects, and validate the algorithm in both a simulation environment as well as on a holographic display setup.
暂无评论