SIGGRAPH 2023 Papers

In this paper, we present a novel method the étendue expansion of near-eye holographic displays through the generation of a sparse eye-box. Conventional holographic near-eye displays have suffered from narrow field of view or narrow eye-box due to the limited étendue supported by a spatial light modulator. We focus on the fact that these displays typically form a dense eye-box, which could be an excessive investment of the limited étendue. By rearranging the eye-box in a sparse manner, the practical étendue can be extended. With a properly designed sparse eye-box shape, it can provide the 3D holographic images and ensure continuous light entrance to the pupil. To create a sparse eye-box, we utilize a lens array as an eyepiece lens. We optimize the spatial light modulator's phase profile for the proposed system and analyze the impact of the use of the lens array eyepiece on the holographic image quality. In particular, we focus on the lens array specification of the lenslet pitch and the focal length, deriving feasible specifications based on our analysis. We experimentally demonstrate a near eye see-through display using the proposed system and verify the étendue expansion.

Large-scale text-to-image generative models have shown their remarkable ability to synthesize diverse, high-quality images. However, directly applying these models for real image editing remains challenging for two reasons. First, it is hard for users to craft a perfect text prompt depicting every visual detail in the input image. Second, while existing models can introduce desirable changes in certain regions, they often dramatically alter the input content and introduce unexpected changes in unwanted regions. In this work, we introduce pix2pix-zero, an image-to-image translation method that can preserve the original image’s content without manual prompting. We first automatically discover editing directions that reflect desired edits in the text embedding space. To preserve the content structure, we propose cross-attention guidance, which aims to retain the cross-attention maps of the input image throughout the diffusion process. Finally, to enable interactive editing, we distill the diffusion model into a fast conditional GAN. We conduct extensive experiments and show that our method outperforms existing and concurrent works for both real and synthetic image editing. In addition, our method does not need additional training for these edits and can directly use the existing pre-trained text-to-image diffusion model.

A word-as-image is a semantic typography technique where a word illustration presents a visualization of the meaning of the word, while also preserving its readability. We present a method to create word-as-image illustrations automatically. This task is highly challenging as it requires semantic understanding of the word and a creative idea of where and how to depict these semantics in a visually pleasing and legible manner. We rely on the remarkable ability of recent large pretrained language-vision models to distill textual concepts visually. We target simple, concise, black-and-white designs that convey the semantics clearly. We deliberately do not change the color or texture of the letters and do not use embellishments. Our method optimizes the outline of each letter to convey the desired concept, guided by a pretrained Stable Diffusion model. We incorporate additional loss terms to ensure the legibility of the text and the preservation of the style of the font. We show high quality and engaging results on numerous examples and compare to alternative techniques. Code and demo will be available at our project page.

In the plane, the winding number is the number of times a curve wraps around a given point. Winding numbers are a basic component of geometric algorithms such as point-in-polygon tests, and their generalization to data with noise or topological errors has proven valuable for geometry processing tasks ranging from surface reconstruction to mesh booleans. However, standard definitions do not immediately apply on surfaces, where not all curves bound regions. We develop a meaningful generalization, starting with the well-known relationship between winding numbers and harmonic functions. By processing the derivatives of such functions, we can robustly filter out components of the input that do not bound any region. Ultimately, our algorithm yields (i) a closed, completed version of the input curves, (ii) integer labels for regions that are meaningfully bounded by these curves, and (iii) the complementary curves that do not bound any region. The main computational cost is solving a standard Poisson equation, or for surfaces with nontrivial topology, a sparse linear program. We also introduce special basis functions to represent singularities that naturally occur at endpoints of open curves.

Grid-free Monte Carlo methods based on the walk on spheres (WoS) algorithm solve fundamental partial differential equations (PDEs) like the Poisson equation without discretizing the problem domain or approximating functions in a finite basis. Such methods hence avoid aliasing in the solution, and evade the many challenges of mesh generation. Yet for problems with complex geometry, practical grid-free methods have been largely limited to basic Dirichlet boundary conditions. We introduce the walk on stars (WoSt) algorithm, which solves linear elliptic PDEs with arbitrary mixed Neumann and Dirichlet boundary conditions. The key insight is that one can efficiently simulate reflecting Brownian motion (which models Neumann conditions) by replacing the balls used by WoS with star-shaped domains. We identify such domains via the closest point on the visibility silhouette, by simply augmenting a standard bounding volume hierarchy with normal information. Overall, WoSt is an easy modification of WoS, and retains the many attractive features of grid-free Monte Carlo methods such as progressive and view-dependent evaluation, trivial parallelization, and sublinear scaling to increasing geometric detail.

Non-line-of-sight (NLOS) imaging methods are capable of reconstructing complex scenes that are not visible to an observer using indirect illumination. However, they assume only third-bounce illumination, so they are currently limited to single-corner configurations, and present limited visibility when imaging surfaces at certain orientations. To reason about and tackle these limitations, we make the key observation that planar diffuse surfaces behave specularly at wavelengths used in the computational wave-based NLOS imaging domain. We call such surfaces virtual mirrors. We leverage this observation to expand the capabilities of NLOS imaging using illumination beyond the third bounce, addressing two problems: imaging single-corner objects at limited visibility angles, and imaging objects hidden behind two corners. To image objects at limited visibility angles, we first analyze the reflections of the known illuminated point on surfaces of the scene as an estimator of the position and orientation of objects with limited visibility. We then image those limited visibility objects by computationally building secondary apertures at other surfaces that observe the target object from a direct visibility perspective. Beyond single-corner NLOS imaging, we exploit the specular behavior of virtual mirrors to image objects hidden behind a second corner by imaging the space behind such virtual mirrors, where the mirror image of objects hidden around two corners is formed. No specular surfaces were involved in the making of this paper.

We present an interactive system to ease the creation of so-called video doodles - videos on which artists insert hand-drawn animations for entertainment or educational purposes. Video doodles are challenging to create because to be convincing, the inserted drawings must appear as if they were part of the captured scene. In particular, the drawings should undergo tracking, perspective deformations and occlusions as they move with respect to the camera and to other objects in the scene - visual effects that are difficult to reproduce with existing 2D video editing software. Our system supports these effects by relying on planar canvases that users position in a 3D scene reconstructed from the video. Furthermore, we present a custom tracking algorithm that allows users to anchor canvases to static or dynamic objects in the scene, such that the canvases move and rotate to follow the position and direction of these objects. When testing our system, novices could create a variety of short animated clips in a dozen of minutes, while professionals praised its speed and ease of use compared to existing tools.

Neural radiance fields (NeRF) have achieved impressive performances in view synthesis by encoding neural representations of a scene. However, NeRFs require hundreds of images per scene to synthesize photo-realistic novel views. Training them on sparse input views leads to overfitting and incorrect scene depth estimation resulting in artifacts in the rendered novel views. Sparse input NeRFs were recently regularized by providing dense depth estimated from pre-trained networks as supervision, to achieve improved performance over sparse depth constraints. However, we find that such depth priors may be inaccurate due to generalization issues. Instead, we hypothesize that the visibility of pixels in different input views can be more reliably estimated to provide dense supervision. In this regard, we compute a visibility prior through the use of plane sweep volumes, which does not require any pre-training. By regularizing the NeRF training with the visibility prior, we successfully train the NeRF with few input views. We reformulate the NeRF to also directly output the visibility of a 3D point from a given viewpoint to reduce the training time with the visibility constraint. On multiple datasets, our model outperforms the competing sparse input NeRF models including those that use learned priors. The source code for our model can be found on our project page: https://nagabhushansn95.github.io/publications/2023/ViP-NeRF.html.

Inspired by the strengths of quadric error metrics initially designed for mesh decimation, we propose a concise mesh reconstruction approach for 3D point clouds. Our approach proceeds by clustering the input points enriched with quadric error metrics, where the generator of each cluster is the optimal 3D point for the sum of its quadric error metrics. This approach favors the placement of generators on sharp features, and tends to equidistribute the error among clusters. We reconstruct the output surface mesh from the adjacency between clusters and a constrained binary solver. We combine our clustering process with an adaptive refinement driven by the error. Compared to prior art, our method avoids dense reconstruction prior to simplification and produces immediately an optimized mesh.

Computation of injective (or inversion-free) maps is a key task in geometry processing, physical simulation, and shape optimization. Despite being a longstanding problem, it remains challenging due to its highly nonconvex and combinatoric nature. We propose computation of variational quasi-harmonic maps to obtain smooth inversion-free maps. Our work is built on a key observation about inversion-free maps: A planar map is a diffeomorphism if and only if it is quasi-harmonic and satisfies a special Cauchy boundary condition. We hence equate the inversion-free mapping problem to an optimal control problem derived from our theoretical result, in which we search in the space of parameters that define an elliptic PDE. We show that this problem can be solved by minimizing within a family of functionals. Similarly, our discretized functionals admit exactly injective maps as the minimizers, empirically producing inversion-free discrete maps of triangle meshes. We design efficient numerical procedures for our problem that prioritize robust convergence paths. Experiments show that on challenging examples our methods can achieve up to orders of magnitude improvement over state-of-the-art, in terms of speed or quality. Moreover, we demonstrate how to optimize a generic energy in our framework while restricting to quasi-harmonic maps.

We present the UrbanBIS benchmark for large-scale 3D urban understanding, supporting practical urban-level semantic and building-level instance segmentation. UrbanBIS comprises six real urban scenes, with 2.5 billion points, covering a vast area of 10.78 km2 and 3,370 buildings, captured by 113,346 views of aerial photogrammetry. Particularly, UrbanBIS provides not only semantic-level annotations on a rich set of urban objects, including buildings, vehicles, vegetation, roads, and bridges, but also instance-level annotations on the buildings. Further, UrbanBIS is the first 3D dataset that introduces fine-grained building sub-categories, considering a wide variety of shapes for different building types. Besides, we propose B-Seg, a building instance segmentation method to establish UrbanBIS. B-Seg adopts an end-to-end framework with a simple yet effective strategy for handling large-scale point clouds. Compared with mainstream methods, B-Seg achieves better accuracy with faster inference speed on UrbanBIS. In addition to the carefully-annotated point clouds, UrbanBIS provides high-resolution aerial-acquisition photos and high-quality large-scale 3D reconstruction models, which shall facilitate a wide range of studies such as multi-view stereo, urban LOD generation, aerial path planning, autonomous navigation, road network extraction, and so on, thus serving as an important platform for many intelligent city applications. UrbanBIS and related code can be downloaded at https://vcc.tech/UrbanBIS.

We propose a novel learning-based approach for robust 3D shape matching. Our method builds upon deep functional maps and can be trained in a fully unsupervised manner. Previous deep functional map methods mainly focus on predicting optimised functional maps alone, and then rely on off-the-shelf post-processing to obtain accurate point-wise maps during inference. However, this two-stage procedure for obtaining point-wise maps often yields sub-optimal performance. In contrast, building upon recent insights about the relation between functional maps and point-wise maps, we propose a novel unsupervised loss to couple the functional maps and point-wise maps, and thereby directly obtain point-wise maps without any post-processing. Our approach obtains accurate correspondences not only for near-isometric shapes, but also for more challenging non-isometric shapes and partial shapes, as well as shapes with different discretisation or topological noise. Using a total of nine diverse datasets, we extensively evaluate the performance and demonstrate that our method substantially outperforms previous state-of-the-art methods, even compared to recent supervised methods. Our code is available at https://github.com/dongliangcao/Unsupervised-Learning-of-Robust-Spectral-Shape-Matching.

Text-driven image generation methods have shown impressive results recently, allowing casual users to generate high quality images by providing textual descriptions. However, similar capabilities for editing existing images are still out of reach. Text-driven image editing methods usually need edit masks, struggle with edits that require significant visual changes and cannot easily keep specific details of the edited portion. In this paper we make the observation that image-generation models can be converted to image-editing models simply by fine-tuning them on a single image. We also show that initializing the stochastic sampler with a noised version of the base image before the sampling and interpolating relevant details from the base image after sampling further increase the quality of the edit operation. Combining these observations, we propose UniTune, a novel image editing method. UniTune gets as input an arbitrary image and a textual edit description, and carries out the edit while maintaining high fidelity to the input image. UniTune does not require additional inputs, like masks or sketches, and can perform multiple edits on the same image without retraining. We test our method using the Imagen model in a range of different use cases. We demonstrate that it is broadly applicable and can perform a surprisingly wide range of expressive editing operations, including those requiring significant visual changes that were previously impossible.

Existing convolutional neural networks have achieved great success in recovering Spatially Varying Bidirectional Surface Reflectance Distribution Function (SVBRDF) maps from a single image. However, they mainly focus on handling low-resolution (e.g., 256 × 256) inputs. Ultra-High Resolution (UHR) material maps are notoriously difficult to acquire by existing networks because (1) finite computational resources set bounds for input receptive fields and output resolutions, and (2) convolutional layers operate locally and lack the ability to capture long-range structural dependencies in UHR images. We propose an implicit neural reflectance model and a divide-and-conquer solution to address these two challenges simultaneously. We first crop a UHR image into low-resolution patches, each of which are processed by a local feature extractor to extract important details. To fully exploit long-range spatial dependency and ensure global coherency, we incorporate a global feature extractor and several coordinate-aware feature assembly modules into our pipeline. The global feature extractor contains several lightweight material vision transformers that have a global receptive field at each scale and have the ability to infer long-term relationships in the material. After decoding globally coherent feature maps assembled by coordinate-aware feature assembly modules, the proposed end-to-end method is able to generate UHR SVBRDF maps from a single image with fine spatial details and consistent global structures.

Visibility computation is a key element in computer graphics applications. More specifically, a from-region potentially visible set (PVS) is an established tool in rendering acceleration, but its high computational cost means a from-region PVS is almost always precomputed. Precomputation restricts the use of PVS to static scenes and leads to high storage cost, in particular, if we need fine-grained regions. For dynamic applications, such as streaming content over a variable-bandwidth network, online PVS computation with configurable region size is required. We address this need with trim regions, a new method for generating from-region PVS for arbitrary scenes in real time. Trim regions perform controlled erosion of object silhouettes in image space, implicitly applying the shrinking theorem known from previous work. Our algorithm is the first that applies automatic shrinking to unconstrained 3D scenes, including non-manifold meshes, and does so in real time using an efficient GPU execution model. We demonstrate that our algorithm generates a tight PVS for complex scenes and outperforms previous online methods for from-viewpoint and from-region PVS. It runs at 60 Hz for realistic game scenes consisting of millions of triangles and computes PVS with a tightness matching or surpassing existing approaches.

Existing devices for measuring material appearance in spatially-varying samples are limited to a single scale, either micro or mesoscopic. This is a practical limitation when the material has a complex multi-scale structure. In this paper, we present a system and methods to digitize materials at two scales, designed to include high-resolution data in spatially-varying representations at larger scales. We design and build a hemispherical light dome able to digitize flat material samples up to 11x11cm. We estimate geometric properties, anisotropic reflectance and transmittance at the microscopic level using polarized directional lighting with a single orthogonal camera. Then, we propagate this structured information to the mesoscale, using a neural network trained with the data acquired by the device and image-to-image translation methods. To maximize the compatibility of our digitization, we leverage standard BSDF models commonly adopted in the industry. Through extensive experiments, we demonstrate the precision of our device and the quality of our digitization process using a set of challenging real-world material samples and validation scenes. Further, we demonstrate the optical resolution and potential of our device for acquiring more complex material representations by capturing microscopic attributes which affect the global appearance: we characterize the properties of textile materials such as the yarn twist or the shape of individual fly-out fibers. We also release the SEDDIDOME dataset of materials, including raw data captured by the machine and optimized parameteres.

Foveated graphics is a promising approach to solving the bandwidth challenges of immersive virtual and augmented reality displays by exploiting the falloff in spatial acuity in the periphery of the visual field. However, the perceptual models used in these applications neglect the effects of higherlevel cognitive processing, namely the allocation of visual attention, and are thus overestimating sensitivity in the periphery in many scenarios. Here, we introduce the first attention-aware model of contrast sensitivity. We conduct user studies to measure contrast sensitivity under different attention distributions and show that sensitivity in the periphery drops significantly when the user is required to allocate attention to the fovea. We motivate the development of future foveation models with another user study and demonstrate that tolerance for foveation in the periphery is significantly higher when the user is concentrating on a task in the fovea. Analysis of our model predicts significant bandwidth savings over those afforded by current models. As such, our work forms the foundation for attention-aware foveated graphics techniques.

Ergonomic efficiency is essential to the mass and prolonged adoption of VR/AR experiences. While VR/AR head-mounted displays unlock users’ natural wide-range head movements during viewing, their neck muscle comfort is inevitably compromised by the added hardware weight. Unfortunately, little quantitative knowledge for understanding and addressing such an issue is available so far.

Computing the intersection between two parametric surfaces (SSI) is one of the most fundamental problems in geometric and solid modeling. Maintaining the SSI topology is critical to its computation robustness. We propose a topology-driven hybrid symbolic-numeric framework to approximate rational parametric surface-surface intersection (SSI) based on a concept of interval algebraic topology analysis (IATA) , which configures within a 4D interval box the SSI topology. We map the SSI topology to an algebraic system's solutions within the framework, classify and enumerate all topological cases as a mixture of four fundamental cases (or their specific sub-cases). Various complicated topological situations are covered, such as cusp points or curves, tangent points (isolated or not) or curves, tiny loops, self-intersections, or their mixtures. The theoretical formulation is also implemented numerically using advanced real solution isolation techniques, and computed within a topology-driven framework which maximally utilizes the advantages of the topology maintenance of algebraic analysis, the robustness of iterative subdivision, and the efficiency of forward marching. The approach demonstrates improved robustness under benchmark topological cases when compared with available open-source and commercial solutions, including IRIT, SISL, and Parasolid.