Sinusoidal neural networks have been shown effective as implicit neural representations (INRs) of low-dimensional signals, due to their smoothness and high representation capacity. However, initializing and training them remain empirical tasks which lack on deeper understanding to guide the learning process. To fill this gap, our work introduces a theoretical framework that explains the capacity property of sinusoidal networks and offers robust control mechanisms for initialization and training. Our analysis is based on a novel amplitude-phase expansion of the sinusoidal multi-layer perceptron, showing how its layer compositions produce a large number of new frequencies expressed as integer combinations of the input frequencies. This relationship can be directly used to initialize the input neurons, as a form of spectral sampling, and to bound the network’s spectrum while training. Our method, referred to as TUNER (TUNing sinusoidal nEtwoRks), greatly improves the stability and convergence of sinusoidal INR training, leading to detailed reconstructions, while preventing overfitting.

Morphing is a long-standing problem in vision and computer graphics, requir- ing a time-dependent warping for feature alignment and a blending for smooth interpolation. Recently, multilayer perceptrons (MLPs) have been explored as implicit neural representations (INRs) for modeling such deformations, due to their meshlessness and differentiability; however, extracting coherent and accurate morphings from standard MLPs typically relies on costly regularizations, which often lead to unstable training and prevent effective feature alignment. To overcome these limitations, we propose FLOWING (FLOW morphING), a framework that recasts warping as the construction of a differential vector flow, naturally ensuring continuity, invertibility, and temporal coherence by encoding structural flow prop- erties directly into the network architectures. This flow-centric approach yields principled and stable transformations, enabling accurate and structure-preserving morphing of both 2D images and 3D shapes. Extensive experiments across a range of applications—including face and image morphing, as well as Gaussian Splatting morphing—show that FLOWING achieves state-of-the-art morphing quality with faster convergence. Code and pretrained models are available in https://schardong.github.io/flowing.

Face morphing is a problem in computer graphics with numerous artistic and forensic applications. It is challenging due to variations in pose, lighting, gender, and ethnicity. This task consists of a warping for feature alignment and a blending for a seamless transition between the warped images. We propose to leverage coord-based neural networks to represent such warpings and blendings of face images. During training, we exploit the smoothness and flexibility of such networks by combining energy functionals employed in classical approaches without discretizations. Additionally, our method is time-dependent, allowing a continuous warping/blending of the images. During morphing inference, we need both direct and inverse transformations of the time-dependent warping. The first (second) is responsible for warping the target (source) image into the source (target) image. Our neural warping stores those maps in a single network dismissing the need for inverting them. The results of our experiments indicate that our method is competitive with both classical and generative models under the lens of image quality and face-morphing detectors. Aesthetically, the resulting images present a seamless blending of diverse faces not yet usual in the literature.

This working-in-progress paper aims to present a three-dimensional reconstruction using aerial images in different environments. The experiments were conducted with aircraft in both external and internal settings, starting with image acquisition, followed by the application of specific photogrammetry software—both commercial and open-source—and concluding with a qualitative evaluation of the results.

This work investigates the use of smooth neural networks for modeling dynamic variations of implicit surfaces under the level set equation (LSE). For this, it extends the representation of neural implicit surfaces to the space-time ℝ3× ℝ, which opens up mechanisms for continuous geometric transformations. Examples include evolving an initial surface towards general vector fields, smoothing and sharpening using the mean curvature equation, and interpolations of initial conditions.The network training considers two constraints. A data term is responsible for fitting the initial condition to the corresponding time instant, usually ℝ3× {0}. Then, a LSE term forces the network to approximate the underlying geometric evolution given by the LSE, without any supervision. The network can also be initialized based on previously trained initial conditions, resulting in faster convergence compared to the standard approach.

Tropical countries face urgent public health challenges regarding epidemic control of Dengue . Since effective vector-control efforts depend on the timing in which public policies take place, there is an enormous demand for accurate prediction tools. In this work, we improve upon a recent approach of coarsely predicting outbreaks in Brazilian urban centers based solely on their yearly climate data. Our methodological advancements encompass a judicious choice of data pre-processing steps and usage of modern computational techniques from signal-processing and manifold learning. Altogether, our results improved earlier prediction accuracy scores from 0.72 to 0.80, solidifying manifold learning on climate data alone as a viable way to make (coarse) dengue outbreak prediction in large urban centers. Ultimately, this approach has the potential of radically simplifying the data required to do outbreak analysis, as municipalities with limited public health funds may not monitor a large number of features needed for more extensive machine learning approaches.

This paper presents a system for immersive visualization of the Classical Non-Euclidean spaces using real-time ray tracing. It exploits the capabilities of the latest generation of GPU's based on the NVIDIA's Turing architecture in order to develop new methods for intuitive exploration of landscapes featuring non-trivial geometry and topology in virtual reality.

VR Kino+Theatre is a media platform that combines theatrical performance with live cinema using virtual reality technology. The platform integrates traditional forms of entertainment (theater and cinema), with advanced interactive media, (virtual reality and gaming). In this way, it solves scalability of audience and presentation familiarity, while providing greater flexibility for innovative formats. The foundations of our solution lies onto three pillars: On the technology front: i) 3D content captured from real data with the help of advanced sensors and machine learning; ii) procedural and real-time physical simulations powered by high-end graphics hardware; iii) distributed systems interconnected by low-latency wireless networks. On the production side: i) unified process, in terms of ubiquitous data access and augmented content generation; ii) collaborative real-time integrated authoring shared by all members of creative teams. On the delivery scenarium: i) diversified media and application options; ii) multiplicity of presentation formats; iii) stratified and complementary fruition allowing to fully explore the content in many forms. The operation of an ecosystem based on these principles entails new roles for producers, performers and participants. As a demonstration of the platform we produced a play “The Tempest”, by William Shakespeare. The project was developed by a multidisciplinary group at IMPA [1].

This paper introduces a new approach to the problem of direct illumination in physically-based rendering of 3D scenes using illumination maps captured from real environments. We developed a system that takes advantage of the best features of the current solutions to the problem: namely, the approximation of illumination maps through directional lights; and stochastic sampling of the light maps. Our framework is flexible and can be used with most rendering programs.

In this work, we introduce a method for resizing a landscape specification, i.e., a vector model containing a set of objects present in a virtual environment. Our goal is to change the landscape dimensions while keeping its overall appearance. Our method is based on the insertion and removal of objects in the specification, followed by some adjustments of the scene adapting the initial model to these changes. Furthermore, this method can be easily extended to use other techniques for spreading objects in the landscape. The adjustment of the scene components consists in performing translations onto the position of the objects based on removal or insertion of paths in the grid created over the scene space, using dynamic programming. This technique is an adaptation of the Seam Carving for vector landscapes specification. This model is simpler than common images, and thus, we can achieve good results using simpler metrics.

In this work, we introduce a method to do object extraction in RGBD images. Our method consists in a depth-based approach which provides an insight into connectedness, proximity and planarity of the scene. We combine the depth and the color in a GraphCut framework to achieve robustness. Specifically, we propose a depth-based seeding which reduces the uncertainty and limitations of the traditional color based seeding. The results of our depth-based seeding were satisfactory and allowed good segmentation results at indoor environments. An extension of our method to do video segmentation using contour graphs is also discussed.

Recent hardware technologies have enabled acquisition of 3D point clouds from real world scenes in real time. A variety of interactive applications with the 3D world can be developed on top of this new technological scenario. However, a main problem that still remains is that most processing techniques for such 3D point clouds are computationally intensive, requiring optimized approaches to handle such images, especially when real time performance is required. As a possible solution, we propose the use of a 3D moving fovea based on a multiresolution technique that processes parts of the acquired scene using multiple levels of resolution. Such approach can be used to identify objects in point clouds with efficient timing. Experiments show that the use of the moving fovea shows a seven fold performance gain in processing time while keeping 91.6% of true recognition rate in comparison with state-of-the-art 3D object recognition methods.

The novelty of our workflow is the end-to-end solution to combine captured panoramas and computer generated elements. This pipeline supports productions specially aiming at spherical displays (e.g., fulldomes). Full panoramas have been used in computer graphics for years, yet their common usage lays on environment lighting and reflection maps for conventional displays. With a keen eye in what may be the next trend in the filmmaking industry, we address the particularities of those productions, proposing a new representation of the space by storing the depth together with the light maps, in a full panoramic light-depth map.

This work presents a new approach for the 3D human facial expressions analysis. Our methodology is based on 2D and 3D wavelet transforms, which are used to estimate multi-scale features from real a face acquired by a 3D scanner. The proposed methodology starts by considering a dataset composed by faces displaying seven different facial expressions. An automatic pre-processing method, adopting an ellipsoidal cropping, is applied to the dataset. Thereafter, the 2D and 3D descriptors are extracted from different scales of wavelet transforms for the purpose of obtaining the facial expression features. The multi-scale features are represented in a multi-variate feature space, which is analysed by the Sequential Forward Floating Selection algorithm using an entropy criterion function to select the subset of features that best represents each facial expression model. The obtained results corroborate the potential of multi-scale feature extraction for analysis of 3D facial expression.

This work presents a method for obtaining perceptually natural panoramic videos, that is, videos composed of wide-angle frames where straight lines and object shapes are preserved. The shape preservation for moving objects has a special treatment that ensures temporal coherence. Energies describing these properties are obtained for the case of a fixed omni-directional camera, and an optimization procedure is presented. Our optimization works per-frame, which makes the method scalable for arbitrarily long scenes.

Abstract The Hermite radial basis functions (HRBF) implicits reconstruct an implicit function which interpolates or approximates scattered multivariate Hermite data (i.e. unstructured points and their corresponding normals). Experiments suggest that HRBF implicits allow the reconstruction of surfaces rich in details and behave better than previous related methods under coarse and/or non‐uniform samplings, even in the presence of close sheets. HRBF implicits theory unifies a recently introduced class of surface reconstruction methods based on radial basis functions (RBF), which incorporate normals directly in their problem formulation. Such class has the advantage of not depending on manufactured offset‐points to ensure existence of a non‐trivial implicit surface RBF interpolant. In fact, we show that HRBF implicits constitute a particular case of Hermite–Birkhoff interpolation with radial basis functions, whose main results we present here. This framework not only allows us to show connections between the present method and others but also enable us to enhance the flexibility of our method by ensuring well‐posedness of an interesting combined interpolation/regularization approach.

Compressive Sensing (CS) is a new paradigm in signal acquisition and compression that has been attracting the interest of the signal compression community. When it comes to image compression applications, it is relevant to estimate the number of bits required to reach a specific image quality. Although several theoretical results regarding the rate-distortion performance of CS have been published recently, there are not many practical image compression results available. The main goal of this paper is to carry out an empirical analysis of the rate-distortion performance of CS in image compression. We analyze issues such as the minimization algorithm used and the transform employed, as well as the trade-off between number of measurements and quantization error. From the experimental results obtained we highlight the potential and limitations of CS when compared to traditional image compression methods.

Abstract The definition of a good view of a 3D scene is highly subjective and strongly depends on both the scene content and the 3D application. Usually, camera placement is performed directly by the user, and that task may be laborious. Existing automatic virtual cameras guide the user by optimizing a single rule, e.g. maximizing the visible silhouette or the projected area. However, the use of a static pre‐defined rule may fail in respecting the user's subjective understanding of the scene. This work introduces intelligent design galleries, a learning approach for subjective problems such as the camera placement. The interaction of the user with a design gallery teaches a statistical learning machine. The trained machine can then imitate the user, either by pre‐selecting good views or by automatically placing the camera. The learning process relies on a Support Vector Machines for classifying views from a collection of descriptors, ranging from 2D image quality to 3D features visibility. Experiments of the automatic camera placement demonstrate that the proposed technique is efficient and handles scenes with occlusion and high depth complexities. This work also includes user validations of the intelligent gallery interface.

An important question about affine skeletons is the existence of differential equations that are related to the “affine distance” and the “area distance” (hence to affine skeletons) in the same way the Eikonal equation is related to the “Euclidean distance” (and the medial axis). We show a surprisingly simple nonlinear second order PDE of Monge–Ampère type that relates to the affine skeletons (and extends the Eikonal equation for the medial axis). We also discuss some consequences and ideas that this new PDE formulation suggests.

Abstract This paper presents a novel segmentation method to assist the rigging of articulated bodies. The method computes a coarse‐to‐fine hierarchy of segments ordered by the level of detail. The results are invariant to deformations, and numerically robust to noise, irregular tessellations, and topological short‐circuits. The segmentation is based on two key ideas. First, it exploits the multiscale properties of the diffusion distance on surfaces, and then it introduces a new definition of medial structures, composing a bijection between medial structures and segments. Our method computes this bijection through a simple and fast iterative approach, and applies it to triangulated meshes.

Abstract Performances of actual mesh compression algorithms vary significantly depending on the type of model it encodes. These methods rely on prior assumptions on the mesh to be efficient, such as regular connectivity, simple topology and similarity between its elements. However, these priors are implicit in usual schemes, harming their suitability for specific models. In particular, connectivity‐driven schemes are difficult to generalize to higher dimensions and to handle topological singularities. GEncode is a new single‐rate, geometry‐driven compression scheme where prior knowledge of the mesh is plugged into the coder in an explicit manner. It encodes meshes of arbitrary dimension without topological restrictions, but can incorporate topological properties, such as manifoldness, to improve the compression ratio. Prior knowledge of the geometry is taken as an input of the algorithm, represented by a function of the local geometry. This suits particularly well for scanned and remeshed models, where exact geometric priors are available. Compression results surfaces and volumes are competitive with existing schemes.

Abstract In computer games and other real‐time graphics applications, the ocean surface is typically modelled as a texture or bump‐mapped plane with simple lighting effects. This paper describes a system for realistically rendering the water surface in real time. Our system can render calm ocean waves with sophisticated lighting effects at 100 fps on a 680 MHz Pentium III with a GeForce 3 graphics card. The wave geometry is represented view‐dependently as a dynamic displacement map with surface detail described by a dynamic bump map. The illumination model includes reflection, refraction and Fresnel effects, which are critical for producing the look and feel of water. Copyright © 2006 John Wiley & Sons, Ltd.

We propose a novel approach for generating video interpolating frames from optical flow information. Our approach is based on solving an affine image-matching equation via the Green's function method. This method provides a pair of filters that, when applied on each frame of an input video signal, allows the generation of a virtual sequence that creates extra frames. Our technique provides smooth reconstruction of video, specifically of sequences created through camera translation or translation of objects in a otherwise static scene.

We describe a new technique that uses active scene illumination to perform foreground-background segmentation and recover partial HDR information. We explore the fact that relative tones can be recovered by varying illumination intensity, without knowing the camera response function. In our approach, the scene is illuminated with an uncalibrated projector and two images of the scene are captured under different illumination conditions. By taking advantage of the fact that the projector can be set up to illuminate only the foreground, we are able to distinguish the foreground from the background. The output of our system is a segmentation mask, together with a image with additional tonal information for the foreground pixels. As an application, we show how to produce spatially variant tone mapped images, where background and foreground receive different treatments. The segmentation and the visualization algorithms are implemented in real-time, and can be used to produce range-enhanced video sequences.

ABSTRACT Deformable models are a powerful tool in both computer graphics and computer vision. The description and implementation of the deformations have to be simultaneously flexible and powerful, otherwise the technique may not satisfy the requirements of all the distinct applications. In this paper, we introduce a new method for the deformable model specification: deformable fields. Deformable fields are conceptually simple, lead to an easy implementation, and are suitable for adaptive models. We apply our new technique to describe an adaptive deformable face, and compare three different adaptation strategies. We show how our technique is suitable to describe different individuals, how to construct a model based on information from a single image, and how it allows the tracking of the deformation parameters over a video sequence.

Abstract This paper proposes the stellar mesh simplification method, a fast implementation of the Four‐Face Cluster (FFC) algorithm. In this method, a probabilistic optimization heuristic substitutes the priority queue of the original method, which results in a 40% faster algorithm with the same order of distortion. It extends naturally to a progressive and/or multiresolution scheme for combinatorial surfaces. This work also presents a simple way to encode the hierarchy of the resulting multiresolution meshes. This work also focuses on important aspects for the development of a practical and robust implementation of this simplification technique, and on the analysis of the influence of the parameters.

Abstract We show how to use affine arithmetic to represent a parametric curve with a strip tree. The required bounding rectangles for pieces of the curve are computed by exploiting the linear correlation information given by affine arithmetic. As an application, we show how to compute approximate distance fields for parametric curves. ACM CSS: I.3.3 Computer Graphics—Curve, surface, solid, and object representations, G.1.2 Numerical Analysis—Approximation of surfaces and contours, G.1.0 Numerical Analysis—Interval arithmetic

We present an approach for decorating surfaces with progressively-variant textures . Unlike a homogeneous texture, a progressively-variant texture can model local texture variations, including the scale, orientation, color, and shape variations of texture elements. We describe techniques for modeling progressively-variant textures in 2D as well as for synthesizing them over surfaces. For 2D texture modeling, our feature-based warping technique allows the user to control the shape variations of texture elements, making it possible to capture complex texture variations such as those seen in animal coat patterns. In addition, our feature-based blending technique can create a smooth transition between two given homogeneous textures, with progressive changes of both shapes and colors of texture elements. For synthesizing textures over surfaces, the biggest challenge is that the synthesized texture elements tend to break apart as they progressively vary. To address this issue, we propose an algorithm based on texton masks, which mark most prominent texture elements in the 2D texture sample. By leveraging the power of texton masks, our algorithm can maintain the integrity of the synthesized texture elements on the target surface.

We propose a new technique that provides a unified framework for modeling the collision dynamics of objects having arbitrary shape and rugosity. The technique assumes that boundary of the object is expressed by elementary segments with some random distribution of perturbation. Collision detection and treatment is done considering this conceptual representation. In order to efficiently detect collisions between irregular objects, we introduce the augmented oriented bounded box tree.

In this paper we introduce a new algorithm for simplification of polygonal meshes. It generates a variable resolution structure called hierarchical 4-K mesh. This structure is a powerful representation for non-uniform level of detail that among other things, allows simple and efficient extraction of conforming meshes.

This paper discusses the problem of motion processing and proposes the use of a mathematical model, which describes a motion signal as a path with variable harmonic components. We show that this model is quite suited to representing a motion path as a periodic part with additional motion texture (noise). This provides a good mathematical formulation to the idea of "motion content". We also describe some mathematical tools that can be used with the model in order to construct different motion paths filters. As an application, we describe an algorithm to change automatically the time duration of a motion. We also point to interesting research directions on motion processing, using the model here introduced.

In this paper we introduce variable resolution 4‐k meshes, a powerful structure for the representation of geometric objects at multiple levels of detail. It combines most properties of other related descriptions with several advantages, such as more flexibility and greater expressive power. The main unique feature of the 4‐k mesh structure lies in its variable resolution capability, which is crucial for adaptive computation. We also give an overview of the different methods for constructing the 4‐k mesh representation, as well as the basic algorithms necessary to incorporate it in modeling and graphics applications.

Direct interactive manipulation of 3D objects is very important in computer graphics. The user interface to deal with this problem poses a difficult problem because of the high number of degrees of freedom involved, associated with the loss of depth in the viewing projection. We introduce a new technique to specify displacement vectors in the three-dimensional space. The technique is built on top of the arcball interface introduced in (Shoemake, 1992). In fact, it constitutes a novel use of the arcball. We also discuss an implementation of the proposed technique in an interactive interface to specify displacement vectors in a widget for surface warping in 3D space.

This paper introduces a unified and general tesselation algorithm for parametric and implicit surfaces. The algorithm produces a hierarchial mesh that is adapted to the surface geometry and has a multiresolution and progressive structure. The representation can be exploited with advantages in several applications.

The recent developments in image-based rendering have enabled a representation of virtual environments based on a simulation of panoramas, which we call virtual panoramas. Current virtual panorama systems do not provide a natural and immersive interaction with the environment. We propose a new system that uses hardware and software components to provide a natural and immersive interaction with virtual panoramas. As part of the system we propose a specific representation for the interactions in a virtual panorama. This representation can be used as a basis for the design of a high-level language for the creation of such environments.

Article Free Access Share on Live paint: painting with procedural multiscale textures Authors: Ken Perlin Media Research Laboratory, Courant Institute of Mathematical Sciences, New York University Media Research Laboratory, Courant Institute of Mathematical Sciences, New York UniversityView Profile , Luiz Velho IMPA, Instituto de Matemática Pura e Aplicada IMPA, Instituto de Matemática Pura e AplicadaView Profile Authors Info & Claims SIGGRAPH '95: Proceedings of the 22nd annual conference on Computer graphics and interactive techniquesSeptember 1995 Pages 153–160https://doi.org/10.1145/218380.218437Published:15 September 1995Publication History 27citation917DownloadsMetricsTotal Citations27Total Downloads917Last 12 Months66Last 6 weeks19 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF

We develop a clustered dithering method that uses stochastic screening and is able to perform an adaptive variation of the cluster size.This makes it possible to achieve optimal rendition of gray shades while preserving image details.The algorithm is an improvement to the dithering with space filling curves method, published in [3].

This paper introduces a new digital halftoning technique that uses space filling curves to generate aperiodic patterns of clustered dots. This method allows the parameterization of the size of pixel clusters, which can vary in one pixel steps. The algorithm unifies, in this way, the dispersed and clustered-dot dithering techniques.

This paper presents a method to generate a polygonal approximation of implicit surfaces. Space partitioning by simplicial decomposition is used to sample the surface, which provides an unambiguous framework for polygon creation. Simplices are adaptively subdivided in order to faithfully represent the surface with a minimum number of elements. The space subdivision is done recursively, without the need of additional data structures to keep track of the process, which usually results in large memory requirements. Furthermore, boundary constraints are imposed on the surface tesselation to overcome the problem of discontinuities usually associated with adaptive methods.