Decomposing 3D assets into material parts is a common task for artists, yet remains a highly manual process. In this work, we introduce Select Any Material (SAMa), a material selection approach for in-the-wild objects in arbitrary 3D representations. Building on SAM2's video prior, we construct a material-centric video dataset that extends it to the material domain. We propose an efficient way to lift the model's 2D predictions to 3D by projecting each view into an intermediary 3D point cloud using depth. Nearestneighbor lookups between any 3D representation and this similarity point cloud allow us to efficiently reconstruct accurate selection masks over objects' surfaces that can be inspected from any view. Our method is multiview-consistent by design, alleviating the need for costly per-asset optimization, and performs optimization-free selection in seconds. SAMa outperforms several strong baselines in selection accuracy and multiview consistency and enables various compelling applications, such as replacing the diffuse-textured materials on a text-to-3D output with PBR materials or selecting and editing materials on NeRFs and 3DGS captures. Project page: https://mfischer-ucl.github.io/sama/.

Abstract Volumetric inverse rendering seeks to recover the optical properties of participating media from images. Existing approaches either rely on differentiable stochastic light transport simulation, which require substantial algorithmic effort, or use simplified models that fail to capture global illumination. We propose a formulation that reconciles physically complete light transport with general‐purpose neural optimization. The optical properties of the medium and the full light field are represented as neural fields and estimated through a joint optimization process. Global illumination is enforced via a residual objective derived from the Radiative Transfer Equation in local differential form, complemented by a volume rendering term along primary viewing rays to mitigate low‐frequency bias. We demonstrate reconstruction of spatially varying, color‐resolved scattering, absorption, and phase function parameters from multi‐view images. Beyond reconstruction, the same framework supports learning generative models of participating media with physical optical properties under global illumination.

We propose a system to optimize parametric designs subject to radiation pressure, i.e., the effect of light on the motion of objects. This is most relevant in the design of spacecraft, where radiation pressure presents the dominant non-conservative forcing mechanism, which is the case beyond approximately 800 km altitude. Despite its importance, the high computational cost of high-fidelity radiation pressure modeling has limited its use in large-scale spacecraft design, optimization, and space situational awareness applications. We enable this by offering three innovations in the simulation, in representation and in optimization: First, a practical computer graphics-inspired Monte-Carlo (MC) simulation of radiation pressure. The simulation is highly parallel, uses importance sampling and next-event estimation to reduce variance and allows simulating an entire family of designs instead of a single spacecraft as in previous work. Second, we introduce neural networks as a representation of forces from design parameters. This neural proxy model, learned from simulations, is inherently differentiable and can query forces orders of magnitude faster than a full MC simulation. Third, and finally, we demonstrate optimizing inverse radiation pressure designs, such as finding geometry, material or operation parameters that minimizes travel time, maximizes proximity given a desired end-point, minimize thruster fuel, trains mission control policies or allocated compute budget in extraterrestrial compute.

We propose a novel generative video model to robustly learn temporal change as a neural Ordinary Differential Equation (ODE) flow with a bilinear objective which combines two aspects: The first is to map from the past into future video frames directly. Previous work has mapped the noise to new frames, a more computationally expensive process. Unfortunately, starting from the previous frame, instead of noise, is more prone to drifting errors. Hence, second, we additionally learn how to remove the accumulated errors as the joint objective by adding noise during training. We demonstrate unconditional video generation in a streaming manner for various video datasets, all at competitive quality compared to a conditional diffusion baseline but with higher speed, i.e., fewer ODE solver steps.

Abstract We demonstrate generating HDR images using the concerted action of multiple black‐box, pre‐trained LDR image diffusion models. Common diffusion models are not HDR as, first, there is no sufficiently large HDR image dataset available to re‐train them, and, second, even if it was, re‐training such models is impossible for most compute budgets. Instead, we seek inspiration from the HDR image capture literature that traditionally fuses sets of LDR images, called “exposure brackets”, to produce a single HDR image. We operate multiple denoising processes to generate multiple LDR brackets that together form a valid HDR result. To this end, we introduce a brackets consistency term into the diffusion process to couple the brackets such that they agree across the exposure range they share. We demonstrate HDR versions of state‐of‐the‐art unconditional and conditional as well as restoration‐type (LDR2HDR) generative modeling.

Real camera footage is subject to noise, motion blur (MB) and depth of field (DoF). In some applications these might be considered distortions to be removed, but in others it is important to model them because it would be ineffective, or interfere with an aesthetic choice, to simply remove them. In augmented reality applications where virtual content is composed into a live video feed, we can model noise, MB and DoF to make the virtual content visually consistent with the video. Existing methods for this typically suffer two main limitations. First, they require a camera calibration step to relate a known calibration target to the specific cameras response. Second, existing work require methods that can be (differentiably) tuned to the calibration, such as slow and specialized neural networks. We propose a method which estimates parameters for noise, MB and DoF instantly, which allows using off-the-shelf real-time simulation methods from e.g., a game engine in compositing augmented content. Our main idea is to unlock both features by showing how to use modern computer vision methods that can remove noise, MB and DoF from the video stream, essentially providing self-calibration. This allows to auto-tune any black-box real-time nose+MB-DoF method to deliver fast and high-fidelity augmentation consistency.

A Neural Radiance Field (NeRF) encodes the specific relation of 3D geometry and appearance of a scene. We here ask the question whether we can transfer the appearance from a source NeRF onto a target 3D geometry in a semantically meaningful way, such that the resulting new NeRF retains the target geometry but has an appearance that is an analogy to the source NeRF. To this end, we generalize classic image analogies from 2D images to NeRFs. We leverage correspondence transfer along semantic affinity that is driven by semantic features from large, pre-trained 2D image models to achieve multi-view consistent appearance transfer. Our method allows exploring the mix-and-match product space of 3D geometry and appearance. We show that our method outperforms traditional stylization-based methods and that a large majority of users prefer our method over several typical baselines. Project page: mfischer-ucl.github.io/nerf_analogies.

Abstract Differentiable rasterization changes the standard formulation of primitive rasterization — by enabling gradient flow from a pixel to its underlying triangles — using distribution functions in different stages of rendering, creating a “soft” version of the original rasterizer. However, choosing the optimal softening function that ensures the best performance and convergence to a desired goal requires trial and error. Previous work has analyzed and compared several combinations of softening. In this work, we take it a step further and, instead of making a combinatorial choice of softening operations, parameterize the continuous space of common softening operations. We study meta‐learning tunable softness functions over a set of inverse rendering tasks (2D and 3D shape, pose and occlusion) so it generalizes to new and unseen differentiable rendering tasks with optimal softness.

Gradient-based optimization is now ubiquitous across graphics, but unfortunately can not be applied to problems with undefined or zero gradients. To circumvent this issue, the loss function can be manually replaced by a "surrogate" that has similar minima but is differentiable. Our proposed framework, ZeroGrads , automates this process by learning a neural approximation of the objective function, which in turn can be used to differentiate through arbitrary black-box graphics pipelines. We train the surrogate on an actively smoothed version of the objective and encourage locality, focusing the surrogate's capacity on what matters at the current training episode. The fitting is performed online, alongside the parameter optimization, and self-supervised, without pre-computed data or pre-trained models. As sampling the objective is expensive (it requires a full rendering or simulator run), we devise an efficient sampling scheme that allows for tractable run-times and competitive performance at little overhead. We demonstrate optimizing diverse non-convex, non-differentiable black-box problems in graphics, such as visibility in rendering, discrete parameter spaces in procedural modelling or optimal control in physics-driven animation. In contrast to other derivative-free algorithms, our approach scales well to higher dimensions, which we demonstrate on problems with up to 35k interlinked variables.

We propose a method to reproduce dynamic appearance textures with space-stationary but time-varying visual statistics. While most previous work decomposes dynamic textures into static appearance and motion, we focus on dynamic appearance that results not from motion but variations of fundamental properties, such as rusting, decaying, melting, and weathering. To this end, we adopt the neural ordinary differential equation (ODE) to learn the underlying dynamics of appearance from a target exemplar. We simulate the ODE in two phases. At the "warm-up" phase, the ODE diffuses a random noise to an initial state. We then constrain the further evolution of this ODE to replicate the evolution of visual feature statistics in the exemplar during the generation phase. The particular innovation of this work is the neural ODE achieving both denoising and evolution for dynamics synthesis, with a proposed temporal training scheme. We study both relightable (BRDF) and non-relightable (RGB) appearance models. For both we introduce new pilot datasets, allowing, for the first time, to study such phenomena: For RGB we provide 22 dynamic textures acquired from free online sources; For BRDFs, we further acquire a dataset of 21 flash-lit videos of time-varying materials, enabled by a simple-to-construct setup. Our experiments show that our method consistently yields realistic and coherent results, whereas prior works falter under pronounced temporal appearance variations. A user study confirms our approach is preferred to previous work for such exemplars.

Current differentiable renderers provide light transport gradients with respect to arbitrary scene parameters. However, the mere existence of these gradients does not guarantee useful update steps in an optimization. Instead, inverse rendering might not converge due to inherent plateaus, i.e., regions of zero gradient, in the objective function. We propose to alleviate this by convolving the high-dimensional rendering function, that maps scene parameters to images, with an additional kernel that blurs the parameter space. We describe two Monte Carlo estimators to compute plateau-reduced gradients efficiently, i.e., with low variance, and show that these translate into net-gains in optimization error and runtime performance. Our approach is a straightforward extension to both black-box and differentiable renderers and enables optimization of problems with intricate light transport, such as caustics or global illumination, that existing differentiable renderers do not converge on. Our code is at github.com/mfischerucl/prdpt.

Abstract We propose a method to accelerate the joint process of physically acquiring and learning neural Bi‐directional Reflectance Distribution Function (BRDF) models. While BRDF learning alone can be accelerated by meta‐learning, acquisition remains slow as it relies on a mechanical process. We show that meta‐learning can be extended to optimize the physical sampling pattern, too. After our method has been meta‐trained for a set of fully‐sampled BRDFs, it is able to quickly train on new BRDFs with up to five orders of magnitude fewer physical acquisition samples at similar quality. Our approach also extends to other linear and non‐linear BRDF models, which we show in an extensive evaluation.

Point patterns are characterized by their density and correlation. While spatial variation of density is well-understood, analysis and synthesis of spatially-varying correlation is an open challenge. No tools are available to intuitively edit such point patterns, primarily due to the lack of a compact representation for spatially varying correlation. We propose a low-dimensional perceptual embedding for point correlations. This embedding can map point patterns to common three-channel raster images, enabling manipulation with off-the-shelf image editing software. To synthesize back point patterns, we propose a novel edge-aware objective that carefully handles sharp variations in density and correlation. The resulting framework allows intuitive and backward-compatible manipulation of point patterns, such as recoloring, relighting to even texture synthesis that have not been available to 2D point pattern design before. Effectiveness of our approach is tested in several user experiments. Code is available at https://github.com/xchhuang/patternshop.

Neural fields are evolving towards a general-purpose continuous representation for visual computing. Yet, despite their numerous appealing properties, they are hardly amenable to signal processing. As a remedy, we present a method to perform general continuous convolutions with general continuous signals such as neural fields. Observing that piecewise polynomial kernels reduce to a sparse set of Dirac deltas after repeated differentiation, we leverage convolution identities and train a repeated integral field to efficiently execute large-scale convolutions. We demonstrate our approach on a variety of data modalities and spatially-varying kernels.

Image-warping, a per-pixel deformation of one image into another, is an essential component in immersive visual experiences such as virtual reality or augmented reality. The primary issue with image warping is disocclusions, where occluded (and hence unknown) parts of the input image would be required to compose the output image. We introduce a new image warping method, Metameric image inpainting - an approach for hole-filling in real-time with foundations in human visual perception. Our method estimates image feature statistics of disoccluded regions from their neighbours. These statistics are inpainted and used to synthesise visuals in real-time that are less noticeable to study participants, particularly in peripheral vision. Our method offers speed improvements over the standard structured image inpainting methods while improving realism over colour-based inpainting such as push-pull. Hence, our work paves the way towards future applications such as depth image-based rendering, 6-DoF 360 rendering, and remote render-streaming.

We introduce 3INGAN, an unconditional 3D generative model trained from 2D images of a single self-similar 3D scene. Such a model can be used to produce 3D “remixes” of a given scene, by mapping spatial latent codes into a 3D volumetric representation, which can subsequently be rendered from arbitrary views using physically based volume rendering. By construction, the generated scenes remain view-consistent across arbitrary camera configurations, without any flickering or spatio-temporal artifacts. During training, we employ a combination of 2D, obtained through differentiable volume tracing, and 3D Generative Adversarial Network (GAN) losses, across multiple scales, enforcing realism on both its 2D renderings and its 3D structure. We show results on semi-stochastic scenes of varying scale and complexity, obtained from real and synthetic sources. We demonstrate, for the first time, the feasibility of learning plausible view-consistent 3D scene variations from a single exemplar scene and provide qualitative and quantitative comparisons against two recent related methods. Code and data for the paper are available at https://geometry.cs.ucl.ac.uk/group_website/projects/2022/3inGAN/.

Scanning Transmission Electron Microscopes (STEMs) acquire 2D images of a 3D sample on the scale of individual cell components. Unfortunately, these 2D images can be too noisy to be fused into a useful 3D structure and facilitating good denoisers is challenging due to the lack of clean-noisy pairs. Additionally, representing detailed 3D structure can be difficult even for clean data when using regular 3D grids. Addressing these two limitations, we suggest a differentiable image formation model for STEM, allowing to learn a joint model of 2D sensor noise in STEM together with an implicit 3D model. We show, that the combination of these models are able to successfully disentangle 3D signal and noise without supervision and outperform at the same time several baselines on synthetic and real data.

Computer-Generated Holography (CGH) offers the potential for genuine, high-quality three-dimensional visuals. However, fulfilling this potential remains a practical challenge due to computational complexity and visual quality issues. We propose a new CGH method that exploits gaze-contingency and perceptual graphics to accelerate the development of practical holographic display systems. Firstly, our method infers the user’s focal depth and generates images only at their focus plane without using any moving parts. Second, the images displayed are metamers; in the user’s peripheral vision, they need only be statistically correct and blend with the fovea seamlessly. Unlike previous methods, our method prioritises and improves foveal visual quality without causing perceptually visible distortions at the periphery. To enable our method, we introduce a novel metameric loss function that robustly compares the statistics of two given images for a known gaze location. In parallel, we implement a model representing the relation between holograms and their image reconstructions. We couple our differentiable loss function and model to metameric varifocal holograms using a stochastic gradient descent solver. We evaluate our method with an actual proof-of-concept holographic display, and we show that our CGH method leads to practical and perceptually three-dimensional image reconstructions.

Abstract We propose a relighting method for outdoor images. Our method mainly focuses on predicting cast shadows in arbitrary novel lighting directions from a single image while also accounting for shading and global effects such the sun light color and clouds. Previous solutions for this problem rely on reconstructing occluder geometry, e. g., using multi‐view stereo, which requires many images of the scene. Instead, in this work we make use of a noisy off‐the‐shelf single‐image depth map estimation as a source of geometry. Whilst this can be a good guide for some lighting effects, the resulting depth map quality is insufficient for directly ray‐tracing the shadows. Addressing this, we propose a learned image space ray‐marching layer that converts the approximate depth map into a deep 3D representation that is fused into occlusion queries using a learned traversal. Our proposed method achieves, for the first time, state‐of‐the‐art relighting results, with only a single image as input. For supplementary material visit our project page at: dgriffiths.uk/outcast .

Abstract Recent advances in neural rendering indicate immense promise for architectures that learn light transport, allowing efficient rendering of global illumination effects once such methods are trained. The training phase of these methods can be seen as a form of pre‐computation, which has a long standing history in Computer Graphics. In particular, Pre‐computed Radiance Transfer (PRT) achieves real‐time rendering by freezing some variables of the scene (geometry, materials) and encoding the distribution of others, allowing interactive rendering at runtime. We adopt the same configuration as PRT – global illumination of static scenes under dynamic environment lighting – and investigate different neural network architectures, inspired by the design principles and theoretical analysis of PRT. We introduce four different architectures, and show that those based on knowledge of light transport models and PRT‐inspired principles improve the quality of global illumination predictions at equal training time and network size, without the need for high‐end ray‐tracing hardware.

There currently exist two main approaches to reproducing visual appearance using Machine Learning (ML): The first is training models that generalize over different instances of a problem, e.g., different images of a dataset. As one-shot approaches, these offer fast inference, but often fall short in quality. The second approach does not train models that generalize across tasks, but rather over-fit a single instance of a problem, e.g., a flash image of a material. These methods offer high quality, but take long to train. We suggest to combine both techniques end-to-end using meta-learning: We over-fit onto a single problem instance in an inner loop, while also learning how to do so efficiently in an outer-loop across many exemplars. To this end, we derive the required formalism that allows applying meta-learning to a wide range of visual appearance reproduction problems: textures, Bidirectional Reflectance Distribution Functions (BRDFs), spatially-varying BRDFs (svBRDFs), illumination or the entire light transport of a scene. The effects of meta-learning parameters on several different aspects of visual appearance are analyzed in our framework, and specific guidance for different tasks is provided. Metappearance enables visual quality that is similar to over-fit approaches in only a fraction of their runtime while keeping the adaptivity of general models.

In this paper we set out to solve the task of 6-DOF 3D object detection from 2D images, where the only supervision is a geometric representation of the objects we aim to find. In doing so, we remove the need for 6-DOF labels (i.e. position, orientation etc.), allowing our network to be trained on unlabeled images in a self-supervised manner. We achieve this through a neural network which learns an explicit scene parameterization which is subsequently passed into a differentiable renderer. We analyze why analysis-by-synthesis-like losses for supervision of 3D scene structure using differentiable rendering is not practical, as it almost always gets stuck in local minima of visual ambiguities. This can be overcome by a novel form of training, where an additional network is employed to steer the optimization itself to explore the entire parameter space i.e. to be curious, and hence, to resolve those ambiguities and find workable minima.

Our goal is to learn a deep network that, given a small number of images of an object of a given category, reconstructs it in 3D. While several recent works have obtained analogous results using synthetic data or assuming the avail-ability of 2D primitives such as keypoints, we are interested in working with challenging real data and with no manual annotations. We thus focus on learning a model from multiple views of a large collection of object instances. We contribute with a new large dataset of object centric videos suitable for training and benchmarking this class of models. We show that existing techniques leveraging meshes, voxels, or implicit surfaces, which work well for reconstructing isolated objects, fail on this challenging data. Finally, we propose a new neural network design, called warp-conditioned ray embedding (WCR), which significantly improves reconstruction while obtaining a detailed implicit representation of the object surface and texture, also compensating for the noise in the initial SfM reconstruction that bootstrapped the learning process. Our evaluation demonstrates performance improvements over several deep monocular reconstruction baselines on existing benchmarks and on our novel dataset. For additional material please visit: https://henzler.github.io/publication/unsupervised_videos/.

Abstract We propose Blue Noise Plots, two‐dimensional dot plots that depict data points of univariate data sets. While often one‐dimensional strip plots are used to depict such data, one of their main problems is visual clutter which results from overlap. To reduce this overlap, jitter plots were introduced, whereby an additional, non‐encoding plot dimension is introduced, along which the data point representing dots are randomly perturbed. Unfortunately, this randomness can suggest non‐existent clusters, and often leads to visually unappealing plots, in which overlap might still occur. To overcome these shortcomings, we introduce Blue Noise Plots where random jitter along the non‐encoding plot dimension is replaced by optimizing all dots to keep a minimum distance in 2D i. e., Blue Noise. We evaluate the effectiveness as well as the aesthetics of Blue Noise Plots through both, a quantitative and a qualitative user study. The Python implementation of Blue Noise Plots is available here.

Abstract Controlled capture of real‐world material appearance yields tabulated sets of highly realistic reflectance data. In practice, however, its high memory footprint requires compressing into a representation that can be used efficiently in rendering while remaining faithful to the original. Previous works in appearance encoding often prioritized one of these requirements at the expense of the other, by either applying high‐fidelity array compression strategies not suited for efficient queries during rendering, or by fitting a compact analytic model that lacks expressiveness. We present a compact neural network‐based representation of BRDF data that combines high‐accuracy reconstruction with efficient practical rendering via built‐in interpolation of reflectance. We encode BRDFs as lightweight networks, and propose a training scheme with adaptive angular sampling, critical for the accurate reconstruction of specular highlights. Additionally, we propose a novel approach to make our representation amenable to importance sampling: rather than inverting the trained networks, we learn to encode them in a more compact embedding that can be mapped to parameters of an analytic BRDF for which importance sampling is known. We evaluate encoding results on isotropic and anisotropic BRDFs from multiple real‐world datasets, and importance sampling performance for isotropic BRDFs mapped to two different analytic models.

Abstract Density estimation plays a crucial role in many data analysis tasks, as it infers a continuous probability density function (PDF) from discrete samples. Thus, it is used in tasks as diverse as analyzing population data, spatial locations in 2D sensor readings, or reconstructing scenes from 3D scans. In this paper, we introduce a learned, data-driven deep density estimation (DDE) to infer PDFs in an accurate and efficient manner, while being independent of domain dimensionality or sample size. Furthermore, we do not require access to the original PDF during estimation, neither in parametric form, nor as priors, or in the form of many samples. This is enabled by training an unstructured convolutional neural network on an infinite stream of synthetic PDFs, as unbound amounts of synthetic training data generalize better across a deck of natural PDFs than any natural finite training data will do. Thus, we hope that our publicly available DDE method will be beneficial in many areas of data analysis, where continuous models are to be estimated from discrete observations.

We learn a latent space for easy capture, consistent interpolation, and efficient reproduction of visual material appearance. When users provide a photo of a stationary natural material captured under flashlight illumination, first it is converted into a latent material code. Then, in the second step, conditioned on the material code, our method produces an infinite and diverse spatial field of BRDF model parameters (diffuse albedo, normals, roughness, specular albedo) that subsequently allows rendering in complex scenes and illuminations, matching the appearance of the input photograph. Technically, we jointly embed all flash images into a latent space using a convolutional encoder, and -conditioned on these latent codes- convert random spatial fields into fields of BRDF parameters using a convolutional neural network (CNN). We condition these BRDF parameters to match the visual characteristics (statistics and spectra of visual features) of the input under matching light. A user study compares our approach favorably to previous work, even those with access to BRDF supervision. Project webpage: https://henzler.github.io/publication/neuralmaterial/.

To peripheral vision, a pair of physically different images can look the same. Such pairs are metamers relative to each other, just as physically-different spectra of light are perceived as the same color. We propose a real-time method to compute such ventral metamers for foveated rendering where, in particular for near-eye displays, the largest part of the framebuffer maps to the periphery. This improves in quality over state-of-the-art foveation methods which blur the periphery. Work in Vision Science has established how peripheral stimuli are ventral metamers if their statistics are similar. Existing methods, however, require a costly optimization process to find such metamers. To this end, we propose a novel type of statistics particularly well-suited for practical real-time rendering: smooth moments of steerable filter responses. These can be extracted from images in time constant in the number of pixels and in parallel over all pixels using a GPU. Further, we show that they can be compressed effectively and transmitted at low bandwidth. Finally, computing realizations of those statistics can again be performed in constant time and in parallel. This enables a new level of quality for foveated applications such as such as remote rendering, level-of-detail and Monte-Carlo denoising. In a user study, we finally show how human task performance increases and foveation artifacts are less suspicious, when using our method compared to common blurring.

We suggest a generative model of 2D and 3D natural textures with diversity, visual fidelity and at high computational efficiency. This is enabled by a family of methods that extend ideas from classic stochastic procedural texturing (Perlin noise) to learned, deep, non-linearities. Our model encodes all exemplars from a diverse set of textures without a need to be re-trained for each exemplar. Applications include texture interpolation, and learning 3D textures from 2D exemplars.

Abstract Convolutional neural networks (CNNs) handle the case where filters extend beyond the image boundary using several heuristics, such as , or padding. These schemes are applied in an ad-hoc fashion and, being weakly related to the image content and oblivious of the target task, result in low output quality at the boundary. In this paper, we propose a simple and effective improvement that learns the boundary handling itself. At training-time, the network is provided with a separate set of boundary filters. At testing-time, we use these filters which have learned to extrapolate features at the boundary in an optimal way for the specific task. Our extensive evaluation, over a wide range of architectural changes (variations of layers, feature channels, or both), shows how the filters result in improved boundary handling. Furthermore, we investigate the efficacy of variations of such boundary filters with respect to convergence speed and accuracy. Finally, we demonstrate an improvement of 5–20% across the board of typical CNN applications (colorization, de-Bayering, optical flow, disparity estimation, and super-resolution). Supplementary material and code can be downloaded from the project page: http://geometry.cs.ucl.ac.uk/projects/2019/investigating-edge/ .

We suggest to represent an X-Field ---a set of 2D images taken across different view, time or illumination conditions, i.e., video, lightfield, reflectance fields or combinations thereof---by learning a neural network (NN) to map their view, time or light coordinates to 2D images. Executing this NN at new coordinates results in joint view, time or light interpolation. The key idea to make this workable is a NN that already knows the "basic tricks" of graphics (lighting, 3D projection, occlusion) in a hard-coded and differentiable form. The NN represents the input to that rendering as an implicit map, that for any view, time, or light coordinate and for any pixel can quantify how it will move if view, time or light coordinates change (Jacobian of pixel position with respect to view, time, illumination, etc.). Our X-Field representation is trained for one scene within minutes, leading to a compact set of trainable parameters and hence real-time navigation in view, time and illumination.

We show that denoising of 3D point clouds can be learned unsupervised, directly from noisy 3D point cloud data only. This is achieved by extending recent ideas from learning of unsupervised image denoisers to unstructured 3D point clouds. Unsupervised image denoisers operate under the assumption that a noisy pixel observation is a random realization of a distribution around a clean pixel value, which allows appropriate learning on this distribution to eventually converge to the correct value. Regrettably, this assumption is not valid for unstructured points: 3D point clouds are subject to total noise, i.e. deviations in all coordinates, with no reliable pixel grid. Thus, an observation can be the realization of an entire manifold of clean 3D points, which makes the quality of a naïve extension of unsupervised image denoisers to 3D point clouds unfortunately only little better than mean filtering. To overcome this, and to enable effective and unsupervised 3D point cloud denoising, we introduce a spatial prior term, that steers converges to the unique closest out of the many possible modes on the manifold. Our results demonstrate unsupervised denoising performance similar to that of supervised learning with clean data when given enough training examples - whereby we do not need any pairs of noisy and clean training data.

We introduce PLATONICGAN to discover the 3D structure of an object class from an unstructured collection of 2D images, i. e., where no relation between photos is known, except that they are showing instances of the same category. The key idea is to train a deep neural network to generate 3D shapes which, when rendered to images, are indistinguishable from ground truth images (for a discriminator) under various camera poses. Discriminating 2D images instead of 3D shapes allows tapping into unstructured 2D photo collections instead of relying on accurated (e. g., aligned, annotated, etc.) 3D data sets. To establish constraints between 2D image observation and their 3D interpretation, we suggest a family of rendering layers that are effectively differentiable. This family includes visual hull, absorption-only (akin to x-ray), and emission-absorption. We can successfully reconstruct 3D shapes from unstructured 2D images and extensively evaluate PLATONICGAN on a range of synthetic and real data sets achieving consistent improvements over baseline methods. We further show that PLATONICGAN can be combined with 3D supervision to improve on and in some cases even surpass the quality of 3D-supervised methods.

We propose a deep representation of appearance, i.e. the relation of color, surface orientation, viewer position, material and illumination. Previous approaches have used deep learning to extract classic appearance representations relating to reflectance model parameters (e.g. Phong) or illumination (e.g. HDR environment maps). We suggest to directly represent appearance itself as a network we call a deep appearance map (DAM). This is a 4D generalization over 2D reflectance maps, which held the view direction fixed. First, we show how a DAM can be learned from images or video frames and later be used to synthesize appearance, given new surface orientations and viewer positions. Second, we demonstrate how another network can be used to map from an image or video frames to a DAM network to reproduce this appearance, without using a lengthy optimization such as stochastic gradient descent (learning-to-learn). Finally, we show the example of an appearance estimation-and-segmentation task, mapping from an image showing multiple materials to multiple deep appearance maps.

Abstract Image metrics predict the perceived per‐pixel difference between a reference image and its degraded (e. g., re‐rendered) version. In several important applications, the reference image is not available and image metrics cannot be applied. We devise a neural network architecture and training procedure that allows predicting the MSE, SSIM or VGG16 image difference from the distorted image alone while the reference is not observed. This is enabled by two insights: The first is to inject sufficiently many un‐distorted natural image patches, which can be found in arbitrary amounts and are known to have no perceivable difference to themselves. This avoids false positives. The second is to balance the learning, where it is carefully made sure that all image errors are equally likely, avoiding false negatives. Surprisingly, we observe that the resulting no‐reference metric, subjectively, can even perform better than the reference‐based one, as it had to become robust against mis‐alignments. We evaluate the effectiveness of our approach in an image‐based rendering context, both quantitatively and qualitatively. Finally, we demonstrate two applications which reduce light field capture time and provide guidance for interactive depth adjustment.

Abstract We suggest a method to directly deep‐learn light transport, i. e., the mapping from a 3D geometry‐illumination‐material configuration to a shaded 2D image. While many previous learning methods have employed 2D convolutional neural networks applied to images, we show for the first time that light transport can be learned directly in 3D. The benefit of 3D over 2D is, that the former can also correctly capture illumination effects related to occluded and/or semi‐transparent geometry. To learn 3D light transport, we represent the 3D scene as an unstructured 3D point cloud, which is later, during rendering, projected to the 2D output image. Thus, we suggest a two‐stage operator comprising a 3D network that first transforms the point cloud into a latent representation, which is later on projected to the 2D output image using a dedicated 3D‐2D network in a second step. We will show that our approach results in improved quality in terms of temporal coherence while retaining most of the computational efficiency of common 2D methods. As a consequence, the proposed two stage‐operator serves as a valuable extension to modern deferred shading approaches.

Abstract Displacement mapping is routinely used to add geometric details in a fast and easy‐to‐control way, both in offline rendering as well as recently in interactive applications such as games. However, it went largely unnoticed (with the exception of McGuire and Whitson [MW08]) that, when applying displacement mapping to a surface with a low‐distortion parametrization, this parametrization is distorted as the geometry was changed by the displacement mapping. Typical resulting artifacts are “rubber band”‐like distortion patterns in areas of strong displacement change where a small isotropic area in texture space is mapped to a large anisotropic area in world space. We describe a fast, fully GPU‐based two‐step procedure to resolve this problem. First, a correction deformation is computed from the displacement map. Second, two variants to apply this correction when computing displacement mapping are proposed. The first variant is backward‐compatible and can resolve the artifact in any rendering pipeline without modifying it and without requiring additional computation at render time, but only works for bijective parametrizations. The second variant works for more general parametrizations, but requires to modify the rendering code and incurs a very small computational overhead.

We suggest a rasterization pipeline tailored towards the needs of HMDs, where latency and field-of-view requirements pose new challenges beyond those of traditional desktop displays. Instead of image warping for low latency, or using multiple passes for foveation, we show how both can be produced directly in a single perceptual rasterization pass. We do this with per-fragment ray-casting. This is enabled by derivations of tight space-time-fovea pixel bounds, introducing just enough flexibility for the requisite geometric tests, but retaining most of the simplicity and efficiency of the traditional rasterizaton pipeline. To produce foveated images, we rasterize to an image with spatially varying pixel density. To compensate for latency, we extend the image formation model to directly produce "rolling" images where the time at each pixel depends on its display location. Our approach overcomes limitations of warping with respect to disocclusions, object motion and view-dependent shading, as well as geometric aliasing artifacts in other foveated rendering techniques. A set of perceptual user studies demonstrates the efficacy of our approach.

Designing point patterns with desired properties can require substantial effort, both in hand-crafting coding and mathematical derivation. Retaining these properties in multiple dimensions or for a substantial number of points can be challenging and computationally expensive. Tackling those two issues, we suggest to automatically generate scalable point patterns from design goals using deep learning. We phrase pattern generation as a deep composition of weighted distance-based unstructured filters. Deep point pattern design means to optimize over the space of all such compositions according to a user-provided point correlation loss , a small program which measures a pattern's fidelity in respect to its spatial or spectral statistics, linear or non-linear (e. g., radial) projections, or any arbitrary combination thereof. Our analysis shows that we can emulate a large set of existing patterns (blue, green, step, projective, stair, etc.-noise), generalize them to countless new combinations in a systematic way and leverage existing error estimation formulations to generate novel point patterns for a user-provided class of integrand functions. Our point patterns scale favorably to multiple dimensions and numbers of points: we demonstrate nearly 10k points in 10-D produced in one second on one GPU. All the resources (source code and the pre-trained networks) can be found at https://sampling.mpi-inf.mpg.de/deepsampling.html.

Faithful manipulation of shape, material, and illumination in 2D Internet images would greatly benefit from a reliable factorization of appearance into material (i.e. diffuse and specular) and illumination (i.e. environment maps). On the one hand, current methods that produce very high fidelity results, typically require controlled settings, expensive devices, or significant manual effort. To the other hand, methods that are automatic and work on 'in the wild' Internet images, often extract only low-frequency lighting or diffuse materials. In this work, we propose to make use of a set of photographs in order to jointly estimate the non-diffuse materials and sharp lighting in an uncontrolled setting. Our key observation is that seeing multiple instances of the same material under different illumination (i.e. environment), and different materials under the same illumination provide valuable constraints that can be exploited to yield a high-quality solution (i.e. specular materials and environment illumination) for all the observed materials and environments. Similar constraints also arise when observing multiple materials in a single environment, or a single material across multiple environments. Technically, we enable this by a novel scalable formulation using parametric mixture models that allows for simultaneous estimation of all materials and illumination directly from a set of (uncontrolled) Internet images. The core of this approach is an optimization procedure that uses two neural networks that are trained on synthetic images to predict good gradients in parametric space given observation of reflected light. We evaluate our method on a range of synthetic and real examples to generate high-quality estimates, qualitatively compare our results against state-of-the-art alternatives via a user study, and demonstrate photo-consistent image manipulation that is otherwise very challenging to achieve.

Abstract As many different 3D volumes could produce the same 2D x‐ray image, inverting this process is challenging. We show that recent deep learning‐based convolutional neural networks can solve this task. As the main challenge in learning is the sheer amount of data created when extending the 2D image into a 3D volume, we suggest firstly to learn a coarse, fixed‐resolution volume which is then fused in a second step with the input x‐ray into a high‐resolution volume. To train and validate our approach we introduce a new dataset that comprises of close to half a million computer‐simulated 2D x‐ray images of 3D volumes scanned from 175 mammalian species. Future applications of our approach include stereoscopic rendering of legacy x‐ray images, re‐rendering of x‐rays including changes of illumination, view pose or geometry. Our evaluation includes comparison to previous tomography work, previous learning methods using our data, a user study and application to a set of real x‐rays.

In this paper, we present a method that estimates reflectance and illumination information from a single image depicting a single-material specular object from a given class under natural illumination. We follow a data-driven, learning-based approach trained on a very large dataset, but in contrast to earlier work we do not assume one or more components (shape, reflectance, or illumination) to be known. We propose a two-step approach, where we first estimate the object's reflectance map, and then further decompose it into reflectance and illumination. For the first step, we introduce a Convolutional Neural Network (CNN) that directly predicts a reflectance map from the input image itself, as well as an indirect scheme that uses additional supervision, first estimating surface orientation and afterwards inferring the reflectance map using a learning-based sparse data interpolation technique. For the second step, we suggest a CNN architecture to reconstruct both Phong reflectance parameters and high-resolution spherical illumination maps from the reflectance map. We also propose new datasets to train these CNNs. We demonstrate the effectiveness of our approach for both steps by extensive quantitative and qualitative evaluation in both synthetic and real data as well as through numerous applications, that show improvements over the state-of-the-art.

Deep learning systems extensively use convolution operations to process input data. Though convolution is clearly defined for structured data such as 2D images or 3D volumes, this is not true for other data types such as sparse point clouds. Previous techniques have developed approximations to convolutions for restricted conditions. Unfortunately, their applicability is limited and cannot be used for general point clouds. We propose an efficient and effective method to learn convolutions for non-uniformly sampled point clouds, as they are obtained with modern acquisition techniques. Learning is enabled by four key novelties: first, representing the convolution kernel itself as a multilayer perceptron; second, phrasing convolution as a Monte Carlo integration problem, third, using this notion to combine information from multiple samplings at different levels; and fourth using Poisson disk sampling as a scalable means of hierarchical point cloud learning. The key idea across all these contributions is to guarantee adequate consideration of the underlying non-uniform sample distribution function from a Monte Carlo perspective. To make the proposed concepts applicable to real-world tasks, we furthermore propose an efficient implementation which significantly reduces the GPU memory required during the training process. By employing our method in hierarchical network architectures we can outperform most of the state-of-the-art networks on established point cloud segmentation, classification and normal estimation benchmarks. Furthermore, in contrast to most existing approaches, we also demonstrate the robustness of our method with respect to sampling variations, even when training with uniformly sampled data only. To support the direct application of these concepts, we provide a ready-to-use TensorFlow implementation of these layers at https://github.com/viscom-ulm/MCCNN.

Simulating combinations of depth-of-field and motion blur is an important factor to cinematic quality in synthetic images but can take long to compute. Splatting the point-spread function (PSF) of every pixel is general and provides high quality, but requires prohibitive compute time. We accelerate this in two steps: In a pre-process we optimize for sparse representations of the Laplacian of all possible PSFs that we call spreadlets. At runtime, spreadlets can be splat efficiently to the Laplacian of an image. Integrating this image produces the final result. Our approach scales faithfully to strong motion and large out-of-focus areas and compares favorably in speed and quality with off-line and interactive approaches. It is applicable to both synthesizing from pinhole as well as reconstructing from stochastic images, with or without layering.

We propose a system to infer binocular disparity from a monocular video stream in real-time. Different from classic reconstruction of physical depth in computer vision, we compute perceptually plausible disparity, that is numerically inaccurate, but results in a very similar overall depth impression with plausible overall layout, sharp edges, fine details and agreement between luminance and disparity. We use several simple monocular cues to estimate disparity maps and confidence maps of low spatial and temporal resolution in real-time. These are complemented by spatially-varying, appearance-dependent and class-specific disparity prior maps, learned from example stereo images. Scene classification selects this prior at runtime. Fusion of prior and cues is done by means of robust MAP inference on a dense spatio-temporal conditional random field with high spatial and temporal resolution. Using normal distributions allows this in constant-time, parallel per-pixel work. We compare our approach to previous 2D-to-3D conversion systems in terms of different metrics, as well as a user study and validate our notion of perceptually plausible disparity.

How much does a single image reveal about the environment it was taken in? In this paper, we investigate how much of that information can be retrieved from a foreground object, combined with the background (i.e. the visible part of the environment). Assuming it is not perfectly diffuse, the foreground object acts as a complexly shaped and far-from-perfect mirror An additional challenge is that its appearance confounds the light coming from the environment with the unknown materials it is made of. We propose a learning-based approach to predict the environment from multiple reflectance maps that are computed from approximate surface normals. The proposed method allows us to jointly model the statistics of environments and material properties. We train our system from synthesized training data, but demonstrate its applicability to real-world data. Interestingly, our analysis shows that the information obtained from objects made out of multiple materials often is complementary and leads to better performance.

Abstract Photographers routinely compose multiple manipulated photos of the same scene into a single image, producing a fidelity difficult to achieve using any individual photo. Alternately, 3D artists set up rendering systems to produce layered images to isolate individual aspects of the light transport, which are composed into the final result in post‐production. Regrettably, these approaches either take considerable time and effort to capture, or remain limited to synthetic scenes. In this paper, we suggest a method to decompose a single image into multiple layers that approximates effects such as shadow, diffuse illumination, albedo, and specular shading. To this end, we extend the idea of intrinsic images along two axes: first, by complementing shading and reflectance with specularity and occlusion, and second, by introducing directional dependence. We do so by training a convolutional neural network (CNN) with synthetic data. Such decompositions can then be manipulated in any off‐the‐shelf image manipulation software and composited back. We demonstrate the effectiveness of our decomposition on synthetic (i. e., rendered) and real data (i. e., photographs), and use them for photo manipulations, which are otherwise impossible to perform based on single images. We provide comparisons with state‐of‐the‐art methods and also evaluate the quality of our decompositions via a user study measuring the effectiveness of the resultant photo retouching setup. Supplementary material and code are available for research use at geometry.cs.ucl.ac.uk/projects/2017/layered-retouching .

Abstract Observing that many visual effects (depth‐of‐field, motion blur, soft shadows, spectral effects) and several sampling modalities (time, stereo or light fields) can be expressed as a sum of many pinhole camera images, we suggest a novel efficient image synthesis framework that exploits coherency among those images. We introduce the notion of “distribution flow” that represents the 2D image deformation in response to changes in the high‐dimensional time‐, lens‐, area light‐, spectral‐, etc. coordinates. Our approach plans the optimal traversal of the distribution space of all required pinhole images, such that starting from one representative root image, which is incrementally changed (warped) in a minimal fashion, pixels move at most by one pixel, if at all. The incremental warping allows extremely simple warping code, typically requiring half a millisecond on an Nvidia Geforce GTX 980Ti GPU per pinhole image. We show, how the bounded sampling does introduce very little errors in comparison to re‐rendering or a common warping‐based solution. Our approach allows efficient previews for arbitrary combinations of distribution effects and imaging modalities with little noise and high visual fidelity.

Texture synthesis has grown into a mature field in computer graphics, allowing the synthesis of naturalistic textures and images from photographic exemplars. Surprisingly little work, however, has been dedicated to synthesizing tileable textures, that is, textures that when laid out in a regular grid of tiles form a homogeneous appearance suitable for use in memory-sensitive real-time graphics applications. One of the key challenges in doing so is that most natural input exemplars exhibit uneven spatial variations that, when tiled, show as repetitive patterns. We propose an approach to synthesize tileable textures while enforcing stationarity properties that effectively mask repetitions while maintaining the unique characteristics of the exemplar. We explore a number of alternative measures for texture stationarity and show how each measure can be integrated into a standard texture synthesis method (PatchMatch) to enforce stationarity at user-controlled scales. We demonstrate the efficacy of our approach using a database of 118 exemplar images, both from publicly available sources as well as new ones captured under uncontrolled conditions, and we quantitatively analyze alternative stationarity measures for their robustness across many test runs using different random seeds. In conclusion, we suggest a novel synthesis approach that employs local histogram matching to reliably turn input photographs of natural surfaces into tiles well suited for artifact-free tiling.

Abstract In computer vision, convolutional neural networks (CNNs) achieve unprecedented performance for inverse problems where RGB pixel appearance is mapped to attributes such as positions, normals or reflectance. In computer graphics, screen space shading has boosted the quality of real‐time rendering, converting the same kind of attributes of a virtual scene back to appearance, enabling effects like ambient occlusion, indirect light, scattering and many more. In this paper we consider the diagonal problem: synthesizing appearance from given per‐pixel attributes using a CNN. The resulting Deep Shading renders screen space effects at competitive quality and speed while not being programmed by human experts but learned from example images.

This paper proposes a pipeline to accurately acquire, efficiently reproduce and intuitively manipulate phosphorescent appearance. In contrast to common appearance models, a model of phosphorescence needs to account for temporal change (decay) and previous illumination (saturation). For reproduction, we propose a rate equation that can be efficiently solved in combination with other illumination in a mixed integro-differential equation system. We describe an acquisition system to measure spectral coefficients of this rate equation for actual materials. Our model is evaluated by comparison to photographs of actual phosphorescent objects. Finally, we propose an artist-friendly interface to control the behavior of phosphorescent materials by specifying spatiotemporal appearance constraints.

Taking an image of an object is at its core a lossy process. The rich information about the three-dimensional structure of the world is flattened to an image plane and decisions such as viewpoint and camera parameters are final and not easily revertible. As a consequence, possibilities of changing viewpoint are limited. Given a single image depicting an object, novel-view synthesis is the task of generating new images that render the object from a different viewpoint than the one given. The main difficulty is to synthesize the parts that are disoccluded; disocclusion occurs when parts of an object are hidden by the object itself under a specific viewpoint. In this work, we show how to improve novel-view synthesis by making use of the correlations observed in 3D models and applying them to new image instances. We propose a technique to use the structural information extracted from a 3D model that matches the image object in terms of viewpoint and shape. For the latter part, we propose an efficient 2D-to-3D alignment method that associates precisely the image appearance with the 3D model geometry with minimal user interaction. Our technique is able to simulate plausible viewpoint changes for a variety of object classes within seconds. Additionally, we show that our synthesized images can be used as additional training data that improves the performance of standard object detectors.

Undoing the image formation process and therefore decomposing appearance into its intrinsic properties is a challenging task due to the under-constrained nature of this inverse problem. While significant progress has been made on inferring shape, materials and illumination from images only, progress in an unconstrained setting is still limited. We propose a convolutional neural architecture to estimate reflectance maps of specular materials in natural lighting conditions. We achieve this in an end-to-end learning formulation that directly predicts a reflectance map from the image itself. We show how to improve estimates by facilitating additional supervision in an indirect scheme that first predicts surface orientation and afterwards predicts the reflectance map by a learning-based sparse data interpolation. In order to analyze performance on this difficult task, we propose a new challenge of Specular MAterials on SHapes with complex IllumiNation (SMASHINg) using both synthetic and real images. Furthermore, we show the application of our method to a range of image editing tasks on real images.

Abstract Light field videos express the entire visual information of an animated scene, but their shear size typically makes capture, processing and display anoff‐lineprocess, i. e., time between initial capture and final display is far from real‐time. In this paper we propose a solution for one of the key bottlenecks in such a processing pipeline, which is a reliable depth reconstruction possibly for many views. This is enabled by a novel correspondence algorithm converting the video streams from a sparse array of off‐the‐shelf cameras into an array of animated depth maps. The algorithm is based on a generalization of the classic multi‐resolution Lucas‐Kanade correspondence algorithm from a pair of images to an entire array. Special inter‐image confidence consolidation allows recovery from unreliable matching in some locations and some views. It can be implemented efficiently in massively parallel hardware, allowing for interactive computations. The resulting depth quality as well as the computation performance compares favorably to other state‐of‐the art light field‐to‐depth approaches, as well as stereo matching techniques. Another outcome of this work is a data set of light field videos that are captured with multiple variants of sparse camera arrays.

Abstract VR headsets and hand‐held devices are not powerful enough to render complex scenes in real‐time. A server can take on the rendering task, but network latency prohibits a good user experience. We present a new image‐based rendering (IBR) architecture for masking the latency. It runs in real‐time even on very weak mobile devices, supports modern game engine graphics, and maintains high visual quality even for large view displacements. We propose a novel server‐sidedual‐viewrepresentation that leverages an optimally‐placed extra view and depth peeling to provide the client with coverage for filling disocclusion holes. This representation is directly rendered in a novel wide‐angle projection with favorable directional parameterization. A new client‐side IBR algorithm uses a pre‐transmitted level‐of‐detail proxy with an encaging simplification and depth‐carving to maintain highly complex geometric detail. We demonstrate our approach with typical VR / mobile gaming applications running on mobile hardware. Our technique compares favorably to competing approaches according to perceptual and numerical comparisons.

Abstract We propose projective blue‐noise patterns that retain their blue‐noise characteristics when undergoing one or multiple projections onto lower dimensional subspaces. These patterns are produced by extending existing methods, such as dart throwing and Lloyd relaxation, and have a range of applications. For numerical integration, our patterns often outperform state‐of‐the‐art stochastic and low‐discrepancy patterns, which have been specifically designed only for this purpose. For image reconstruction, our method outperforms traditional blue‐noise sampling when the variation in the signal is concentrated along one dimension. Finally, we use our patterns to distribute primitives uniformly in 3D space such that their 2D projections retain a blue‐noise distribution.

Our aim is to give users real-time free-viewpoint rendering of real indoor scenes, captured with off-the-shelf equipment such as a high-quality color camera and a commodity depth sensor. Image-based Rendering (IBR) can provide the realistic imagery required at real-time speed. For indoor scenes however, two challenges are especially prominent. First, the reconstructed 3D geometry must be compact, but faithful enough to respect occlusion relationships when viewed up close. Second, man-made materials call for view-dependent texturing, but using too many input photographs reduces performance. We customize a typical RGB-D 3D surface reconstruction pipeline to produce a coarse global 3D surface, and local, per-view geometry for each input image. Our tiled IBR preserves quality by economizing on the expected contributions that entire groups of input pixels make to a final image. The two components are designed to work together, giving real-time performance, while hardly sacrificing quality. Testing on a variety of challenging scenes shows that our inside-out IBR scales favorably with the number of input images.

Binocular disparity is the main depth cue that makes stereoscopic images appear 3D. However, in many scenarios, the range of depth that can be reproduced by this cue is greatly limited and typically fixed due to constraints imposed by displays. For example, due to the low angular resolution of current automultiscopic screens, they can only reproduce a shallow depth range. In this work, we study the motion parallax cue, which is a relatively strong depth cue, and can be freely reproduced even on a 2D screen without any limits. We exploit the fact that in many practical scenarios, motion parallax provides sufficiently strong depth information that the presence of binocular depth cues can be reduced through aggressive disparity compression. To assess the strength of the effect we conduct psycho-visual experiments that measure the influence of motion parallax on depth perception and relate it to the depth resulting from binocular disparity. Based on the measurements, we propose a joint disparity-parallax computational model that predicts apparent depth resulting from both cues. We demonstrate how this model can be applied in the context of stereo and multiscopic image processing, and propose new disparity manipulation techniques, which first quantify depth obtained from motion parallax, and then adjust binocular disparity information accordingly. This allows us to manipulate the disparity signal according to the strength of motion parallax to improve the overall depth reproduction. This technique is validated in additional experiments.

Different from classic reconstruction of physical depth in computer vision, depth for 2D-to-3D stereo conversion is assigned by humans using semi-automatic painting interfaces and, consequently, is often dramatically wrong. Here we seek to better understand why it still does not fail to convey a sensation of depth. To this end, four typical disparity distortions resulting from manual 2D-to-3D stereo conversion are analyzed: i) smooth remapping, ii) spatial smoothness, iii) motion-compensated, temporal smoothness, and iv) completeness. A perceptual experiment is conducted to quantify the impact of each distortion on the plausibility of the 3D impression relative to a reference without distortion. Close-to-natural videos with known depth were distorted in one of the four above-mentioned aspects and subjects had to indicate if the distortion still allows for a plausible 3D effect. The smallest amounts of distortion that result in a significant rejection suggests a conservative upper bound on the quality requirement of 2D-to-3D conversion.

Abstract Providing multiple meanings in a single piece of art has always been intriguing to both artists and observers. We present Purkinje images, which have different interpretations depending on the luminance adaptation of the observer. Finding such images is an optimization that minimizes the sum of the distance to one reference image in photopic conditions and the distance to another reference image in scotopic conditions. To model the shift of image perception between day and night vision, we decompose the input images into a Laplacian pyramid. Distances under different observation conditions in this representation are independent between pyramid levels and pixel positions and become matrix multiplications. The optimal pixel colour can be found by inverting a small, per‐pixel linear system in real time on a GPU. Finally, two user studies analyze our results in terms of the recognition performance and fidelity with respect to the reference images.

Abstract When human luminance perception operates close to its absolute threshold, i. e., the lowest perceivable absolute values, appearance changes substantially compared to common photopic or scotopic vision. In particular, most observers report perceiving temporally‐varying noise. Two reasons are physiologically plausible; quantum noise (due to the low absolute number of photons) and spontaneous photochemical reactions. Previously, static noise with a normal distribution and no account for absolute values was combined with blue hue shift and blur to simulate scotopic appearance on a photopic display for movies and interactive applications (e.g., games). We present a computational model to reproduce the specific distribution and dynamics of “scotopic noise” for specific absolute values. It automatically introduces a perceptually‐calibrated amount of noise for a specific luminance level and supports animated imagery. Our simulation runs in milliseconds at HD resolution using graphics hardware and favorably compares to simpler alternatives in a perceptual experiment.

Abstract We propose a system to restrict the manipulation of shape and appearance in an image to a valid subspace which we learn from a collection of exemplar images. To this end, we automatically co‐align a collection of images and learn a subspace model of shape and appearance using principal components. As finding perfect image correspondences for general images is not feasible, we build an approximate partial alignment and improve bad alignments leveraging other, more successful alignments. Our system allows the user to change appearance and shape in real‐time and the result is “projected” onto the subspace of meaningful changes. The change in appearance and shape can either be locked or performed independently. Additional applications include suggestion of alternative shapes or appearance.

Color selection is required in many computer graphics applications, but can be tedious, as 1D or 2D user interfaces are employed to navigate in a 3D color space. Until now the problem was considered a question of designing general color spaces with meaningful (e.g., perceptual) parameters. In this work, we show how color selection usability improves by applying 1D or 2D color manifolds that predict the most likely change of color in a specific context. A typical use-case is manipulating the color of a banana; instead of presenting a 2D+1D RGB, CIE Lab, or HSV widget, our approach presents a simple 1D slider that captures the most likely change for this context. Technically, for each context, we learn a lower-dimensional manifold with varying density from labeled Internet examples. We demonstrate the increase in task performance of color selection in a user study.

We propose a technique to use the structural informa- tion extracted from a set of 3D models of an object class to improve novel-view synthesis for images showing unknown instances of this class. These novel views can be used to "amplify" training image collections that typically contain only a low number of views or lack certain classes of views entirely (e. g. top views). We extract the correlation of position, normal, re- flectance and appearance from computer-generated images of a few exemplars and use this information to infer new appearance for new instances. We show that our approach can improve performance of state-of-the-art detectors using real-world training data. Additional applications include guided versions of inpainting, 2D-to-3D conversion, super- resolution and non-local smoothing.

Computing shading such as ambient occlusion (AO), subsurface scattering (SSS) or indirect light (GI) in screen space has recently received a lot of attention. While being efficient to compute, screen space methods have several key limitations such as occlusions, culling, under-sampling of oblique geometry and locality of the transport. In this work we propose a deep screen space to overcome all these problems while retaining computational efficiency. Instead of projecting, culling, shading, rasterizing and resolving occlusions of primitives using a z-buffer, we adaptively tessellate them into surfels proportional to the primitive's projected size, which are optionally shaded and stored on-GPU as an unstructured surfel cloud. Objects closer to the camera receive more details, like in classic framebuffers, but are not affected by occlusions or viewing angle. This surfel cloud can then be used to compute shading. Instead of gathering, we propose to use splatting to a multi-resolution interleaved framebuffer. This allows to exchange detailed shading between pixels close to a surfel and approximate shading between pixels distant to a surfel.

Abstract Presenting stereoscopic content on 3D displays is a challenging task, usually requiring manual adjustments. A number of techniques have been developed to aid this process, but they account for binocular disparity of surfaces that are diffuse and opaque only. However, combinations of transparent as well as specular materials are common in the real and virtual worlds, and pose a significant problem. For example, excessive disparities can be created which cannot be fused by the observer. Also, multiple stereo interpretations become possible, e. g., for glass, that both reflects and refracts, which may confuse the observer and result in poor 3D experience. In this work, we propose an efficient method for analyzing and controlling disparities in computer‐generated images of such scenes where surface positions and a layer decomposition are available. Instead of assuming a single per‐pixel disparity value, we estimate all possibly perceived disparities at each image location. Based on this representation, we define an optimization to find the best per‐pixel camera parameters, assuring that all disparities can be easily fused by a human. A preliminary perceptual study indicates, that our approach combines comfortable viewing with realistic depiction of typical specular scenes.

Abstract Light refracted by a dispersive interface leads to beautifully colored patterns that can be rendered faithfully with spectral Monte‐Carlo methods. Regrettably, results often suffer from chromatic noise or banding, requiring high sampling rates and large amounts of memory compared to renderers operating in some trichromatic color space. Addressing this issue, we introduce spectral ray differentials, which describe the change of light direction with respect to changes in the spectrum. In analogy with the classic ray and photon differentials, this information can be used for filtering in the spectral domain. Effectiveness of our approach is demonstrated by filtering for offline spectral light and path tracing as well as for an interactive GPU photon mapper based on splatting. Our results show considerably less chromatic noise and spatial aliasing while retaining good visual similarity to reference solutions with negligible overhead in the order of milliseconds.

Several approaches attempt to reproduce the appearance of a scotopic low-light night scene on a photopic display (“day-for-night”) by introducing color desaturation, loss of acuity, and the Purkinje shift toward blue colors. We argue that faithful stereo reproduction of night scenes on photopic stereo displays requires manipulation of not only color but also binocular disparity. To this end, we performed a psychophysics experiment to devise a model of disparity at scotopic luminance levels. Using this model, we can match binocular disparity of a scotopic stereo content displayed on a photopic monitor to the disparity that would be perceived if the scene was actually scotopic. The model allows for real-time computation of common stereo content as found in interactive applications such as simulators or computer games.

Monte Carlo techniques for light transport simulation rely on importance sampling when constructing light transport paths. Previous work has shown that suitable sampling distributions can be recovered from particles distributed in the scene prior to rendering. We propose to represent the distributions by a parametric mixture model trained in an on-line (i.e. progressive) manner from a potentially infinite stream of particles. This enables recovering good sampling distributions in scenes with complex lighting, where the necessary number of particles may exceed available memory. Using these distributions for sampling scattering directions and light emission significantly improves the performance of state-of-the-art light transport simulation algorithms when dealing with complex lighting.

Abstract Beyond the careful design of stereo acquisition equipment and rendering algorithms, disparity post‐processing has recently received much attention, where one of the key tasks is to compress the originally large disparity range to avoid viewing discomfort. The perception of dynamic stereo content however, relies on reproducing the full disparity‐time volume that a scene point undergoes in motion. This volume can be strongly distorted in manipulation, which is only concerned with changing disparity at one instant in time, even if the temporal coherence of that change is maintained. We propose an optimization to preserve stereo motion of content that was subject to an arbitrary disparity manipulation, based on a perceptual model of temporal disparity changes. Furthermore, we introduce a novel 3D warping technique to create stereo image pairs that conform to this optimized disparity map. The paper concludes with perceptual studies of motion reproduction quality and task performance in a simple game, showing how our optimization can achieve both viewing comfort and faithful stereo motion.

Abstract This work addresses the challenge of intuitive appearance editing in scenes with complex geometric layout and complex, spatially‐varying indirect lighting. In contrast to previous work, that aimed to edit surface reflectance, our system allows a user to freely manipulate the surface light field. It then finds the best surface reflectance that “explains” the surface light field manipulation. Instead of classic L2fitting of reflectance to a combination of incoming and exitant illumination, our system infers a sparse L0change of shading parameters instead. Consequently, our system does not require “diffuse” or “glossiness” brushes or any such understanding of the underlying reflectance parametrization. Instead, it infers reflectance changes from scribbles made by a single simple color brush tool alone: Drawing a highlight will increase Phong specular; blurring a mirror reflection will decrease glossiness; etc. A sparse‐solver framework operating on a novel point‐based, pre‐convolved lighting representation in combination with screen‐space edit upsampling allows to perform editing interactively on a GPU.

We propose an approach to "pack" a set of two-dimensional graphical primitives into a spatial layout that follows artistic goals. We formalize this process as projecting from a high-dimensional feature space into a 2D layout. Our system does not expose the control of this projection to the user in form of sliders or similar interfaces. Instead, we infer the desired layout of all primitives from interactive placement of a small subset of example primitives. To produce a pleasant distribution of primitives with spatial extend, we propose a novel generalization of Centroidal Voronoi Tesselation which equalizes the distances between boundaries of nearby primitives. Compared to previous primitive distribution approaches our GPU implementation achieves both better fidelity and asymptotically higher speed. A user study evaluates the system's usability.

Abstract We present a system that allows new shapes to be created by blending between shapes taken from a database. We treat the shape as a composition of parts; blending is performed by recombining parts from different shapes according to constraints deduced by shape analysis. The analysis involves shape segmentation, contact analysis, and symmetry detection. The system can be used to rapidly instantiate new models that have similar symmetry and adjacency structure to the database shapes, yet vary in appearance.

Abstract This work proposes a technique to transfer the material style or mood from a guide source such as an image or video onto a target 3D scene. It formulates the problem as a combinatorial optimization of assigning discrete materials extracted from the guide source to discrete objects in the target 3D scene. The assignment is optimized to fulfill multiple goals: overall image mood based on several image statistics; spatial material organization and grouping as well as geometric similarity between objects that were assigned to similar materials. To be able to use common uncalibrated images and videos with unknown geometry and lighting as guides, a material estimation derives perceptually plausible reflectance, specularity, glossiness, and texture. Finally, results produced by our method are compared to manual material assignments in a perceptual study.

Abstract Size matters. Human perception most naturally relates relative extent, area or volume to importance, nearness and weight. Reversely, conveying importance of something by depicting it at a different size is a classic artistic principle, in particular when importance varies across a domain. One striking example is the neuronal homunculus; a human figure where the size of each body part is proportional to the neural density on that part. In this work we propose an approach which changes local size of a 2D image or 3D surface and, at the same time, minimizes distortion, prevails smoothness, and, most importantly, avoids fold‐overs (collisions). We employ a parallel, two‐stage optimization process, that scales the shape non‐uniformly according to an interactively‐defined importance map and then solves for a nearby, self‐intersection‐free configuration. The results include an interactive 3D‐rendered version of the classic sensorical homunculus but also a range of images and surfaces with different importance maps.

Abstract The interaction of light and matter in the world surrounding us is of striking complexity and beauty. Since the very beginning of computer graphics, adequate modelling of these processes and efficient computation is an intensively studied research topic and still not a solved problem. The inherent complexity stems from the underlying physical processes as well as the global nature of the interactions that let light travel within a scene. This paper reviews the state of the art in interactive global illumination (GI) computation, i.e., methods that generate an image of a virtual scene in less than 1 s with an as exact as possible, or plausible, solution to the light transport. Additionally, the theoretical background and attempts to classify the broad field of methods are described. The strengths and weaknesses of different approaches, when applied to the different visual phenomena, arising from light interaction are compared and discussed. Finally, the paper concludes by highlighting design patterns for interactive GI and a list of open problems.

Abstract Afterimages are optical illusions, particularly well perceived when fixating an image for an extended period of time and then looking at a neutral background, where an inverted copy of the original stimulus appears. The full mechanism that produces the perceived specific colors and shapes is complex and not entirely understood, but most of the important attributes can be well explained by bleaching of retinal photoreceptors (retinal kinetics). We propose a model to compute afterimages that allows us to simulate their temporal, color and time‐frequency behavior. Using this model, high dynamic range (HDR) content can be processed to add realistic afterimages to low dynamic range (LDR) media. Hereby, our approach helps in conveying the original source's luminance and contrast. It can be applied in real‐time on full‐HD HDR content using standard graphics hardware. Finally, our approach is validated in a perceptual study.

Abstract The incident indirect light over a range of image pixels is often coherent. Two common approaches to exploit this inter‐pixel coherence to improve rendering performance are Irradiance Caching and Radiance Caching. Both compute incident indirect light only for a small subset of pixels (the cache), and later interpolate between pixels. Irradiance Caching uses scalar values that can be interpolated efficiently, but cannot account for shading variations caused by normal and reflectance variation between cache items. Radiance Caching maintains directional information, e.g., to allow highlights between cache items, but at the cost of storing and evaluating a Spherical Harmonics (SH) function per pixel. The arithmetic and bandwidth cost for this evaluation is linear in the number of coefficients and can be substantial. In this paper, we propose a method to replace it by an efficient per‐cache item pre‐filtering based on MIP maps — such as previously done for environment maps — leading to a single constant‐time lookup per pixel. Additionally, per‐cache item geometry statistics stored in distance‐MIP maps are used to improve the quality of each pixel's lookup. Our approximate interactive global illumination approach is an order of magnitude faster than Radiance Caching with Phong BRDFs and can be combined with Monte Carlo‐raytracing, Point‐based Global Illumination or Instant Radiosity.

Binocular disparity is one of the most important depth cues used by the human visual system. Recently developed stereo-perception models allow us to successfully manipulate disparity in order to improve viewing comfort, depth discrimination as well as stereo content compression and display. Nonetheless, all existing models neglect the substantial influence of luminance on stereo perception. Our work is the first to account for the interplay of luminance contrast (magnitude/frequency) and disparity and our model predicts the human response to complex stereo-luminance images. Besides improving existing disparity-model applications (e.g., difference metrics or compression), our approach offers new possibilities, such as joint luminance contrast and disparity manipulation or the optimization of auto-stereoscopic content. We validate our results in a user study, which also reveals the advantage of considering luminance contrast and its significant impact on disparity manipulation techniques.

The material found on 3D objects and their parts in our everyday surroundings is highly correlated with the geometric shape of the parts and their relation to other parts of the same object. This work proposes to model this context-dependent correlation by learning it from a database containing several hundreds of objects and their materials. Given a part-based 3D object without materials, the learned model can be used to fully automatically assign plausible material parameters, including diffuse color, specularity, gloss, and transparency. Further, we propose a user interface that provides material suggestions. This user-interface can be used, for example, to refine the automatic suggestion. Once a refinement has been made, the model incorporates this information, and the automatic assignment is incrementally improved. Results are given for objects with different numbers of parts and with different topological complexity. A user study validates that our method significantly simplifies and accelerates the material assignment task compared to other approaches.

Human stereo perception of glossy materials is substantially different from the perception of diffuse surfaces: A single point on a diffuse object appears the same for both eyes, whereas it appears different to both eyes on a specular object. As highlights are blurry reflections of light sources they have depth themselves, which is different from the depth of the reflecting surface. We call this difference in depth impression the "highlight disparity". Due to artistic motivation, for technical reasons, or because of incomplete data, highlights often have to be depicted on-surface, without any disparity. However, it has been shown that a lack of disparity decreases the perceived glossiness and authenticity of a material. To remedy this contradiction, our work introduces a technique for depiction of glossy materials, which improves over simple on-surface highlights, and avoids the problems of physical highlights. Our technique is computationally simple, can be easily integrated in an existing (GPU) shading system, and allows for local and interactive artistic control.

Abstract Level‐of‐Detail structures are a key component for scalable rendering. Built from raw 3D data, these structures are often defined as Bounding Volume Hierarchies, providing coarse‐to‐fine adaptive approximations that are well‐adapted formany‐viewrasterization. Here, the total number of pixels in each view is usually low, while the cost of choosing the appropriate LoD for each view is high. This task represents a challenge for existing GPU algorithms. We propose ManyLoDs, a new GPU algorithm to efficiently compute many LoDs from a Bounding Volume Hierarchy in parallel by balancing the workload within and among LoDs. Our approach is not specific to a particular rendering technique, can be used on lazy representations such as polygon soups, and can handle dynamic scenes. We apply our method to various many‐view rasterization applications, including Instant Radiosity, Point‐Based Global Illumination, and reflection/refraction mapping. For each of these, we achieve real‐time performance in complex scenes at high resolutions.

Abstract We propose an algorithm to compute interactive indirect illumination in dynamic scenes containing millions of triangles. It makes use of virtual point lights (VPL) to compute bounced illumination and a point‐based scene representation to query indirect visibility, similar to Imperfect Shadow Maps (ISM). To ensure a high fidelity of indirect light and shadows, our solution is made view‐adaptive by means of two orthogonal improvements: First, the VPL distribution is chosen to provide more detail, that is, more dense VPL sampling, where these contribute most to the current view. Second, the scene representation for indirect visibility is adapted to ensure geometric detail where it affects indirect shadows in the current view.

Abstract Finding the best makeup for a given human face is an art in its own right. Experienced makeup artists train for years to be skilled enough to propose a best‐fit makeup for an individual. In this work we propose a system that automates this task. We acquired the appearance of 56 human faces, both without and with professional makeup. To this end, we use a controlled‐light setup, which allows to capture detailed facial appearance information, such as diffuse reflectance, normals, subsurface‐scattering, specularity, or glossiness. A 3D morphable face model is used to obtain 3D positional information and to register all faces into a common parameterization. We then define makeup to be the change of facial appearance and use the acquired database to find a mapping from the space of human facial appearance to makeup. Our main application is to use this mapping to suggest the best‐fit makeup for novel faces that are not in the database. Further applications are makeup transfer, automatic rating of makeup, makeup‐training, or makeup‐exaggeration. As our makeup representation captures a change in reflectance and scattering, it allows us to synthesize faces with makeup in novel 3D views and novel lighting with high realism. The effectiveness of our approach is further validated in a user‐study.

Binocular disparity is an important cue for the human visual system to recognize spatial layout, both in reality and simulated virtual worlds. This paper introduces a perceptual model of disparity for computer graphics that is used to define a metric to compare a stereo image to an alternative stereo image and to estimate the magnitude of the perceived disparity change. Our model can be used to assess the effect of disparity to control the level of undesirable distortions or enhancements (introduced on purpose). A number of psycho-visual experiments are conducted to quantify the mutual effect of disparity magnitude and frequency to derive the model. Besides difference prediction, other applications include compression, and re-targeting. We also present novel applications in form of hybrid stereo images and backward-compatible stereo. The latter minimizes disparity in order to convey a stereo impression if special equipment is used but produces images that appear almost ordinary to the naked eye. The validity of our model and difference metric is again confirmed in a study.

Color perception is recognized to vary with surrounding spatial structure, but the impact of edge smoothness on color has not been studied in color appearance modeling. In this work, we study the appearance of color under different degrees of edge smoothness. A psychophysical experiment was conducted to quantify the change in perceived lightness, colorfulness, and hue with respect to edge smoothness. We confirm that color appearance, in particular lightness, changes noticeably with increased smoothness. Based on our experimental data, we have developed a computational model that predicts this appearance change. The model can be integrated into existing color appearance models. We demonstrate the applicability of our model on a number of examples.

Abstract High‐refresh‐rate displays (e. g., 120 Hz) have recently become available on the consumer market and quickly gain on popularity. One of their aims is to reduce the perceived blur created by moving objects that are tracked by the human eye. However, an improvement is only achieved if the video stream is produced at the same high refresh rate (i. e. 120 Hz). Some devices, such as LCD TVs, solve this problem by converting low‐refresh‐rate content (i. e. 50 Hz PAL) into a higher temporal resolution (i. e. 200 Hz) based on two‐dimensional optical flow. In our approach, we will show how rendered three‐dimensional images produced by recent graphics hardware can be up‐sampled more efficiently resulting in higher quality at the same time. Our algorithm relies on several perceptual findings and preserves the naturalness of the original sequence. A psychophysical study validates our approach and illustrates that temporally up‐sampled video streams are preferred over the standard low‐rate input by the majority of users. We show that our solution improves task performance on high‐refresh‐rate displays.

Limited spatial resolution of current displays makes the depiction of very fine spatial details difficult. This work proposes a novel method applied to moving images that takes into account the human visual system and leads to an improved perception of such details. To this end, we display images rapidly varying over time along a given trajectory on a high refresh rate display. Due to the retinal integration time the information is fused and yields apparent super-resolution pixels on a conventional-resolution display. We discuss how to find optimal temporal pixel variations based on linear eye-movement and image content and extend our solution to arbitrary trajectories. This step involves an efficient method to predict and successfully treat potentially visible flickering. Finally, we evaluate the resolution enhancement in a perceptual study that shows that significant improvements can be achieved both for computer generated images and photographs.

We present an interactive system for the artistic control of visual phenomena visible on surfaces. Our method allows the user to intuitively reposition shadows, caustics, and indirect illumination using a simple click-and-drag user interface working directly on surfaces. In contrast to previous approaches, the positions of the lights or objects in the scene remain unchanged, enabling localized edits of individual shading components. Our method facilitates the editing by computing a mapping from one surface location to another. Based on this mapping, we can not only edit shadows, caustics, and indirect illumination but also other surface properties, such as color or texture, in a unified way. This is achieved using an intuitive user-interface that allows the user to specify position constraints with drag-and-drop or sketching operations directly on the surface. Our approach requires no explicit surface parametrization and handles scenes with arbitrary topology. We demonstrate the applicability of the approach to interactive editing of shadows, reflections, refractions, textures, caustics, and diffuse indirect light. The effectiveness of the system to achieve an artistic goal is evaluated by a user study.

Physically plausible illumination at real-time framerates is often achieved using approximations. One popular example is ambient occlusion (AO), for which very simple and efficient implementations are used extensively in production. Recent methods approximate AO between nearby geometry in screen space (SSAO). The key observation described in this paper is, that screen-space occlusion methods can be used to compute many more types of effects than just occlusion, such as directional shadows and indirect color bleeding. The proposed generalization has only a small overhead compared to classic SSAO, approximates direct and one-bounce light transport in screen space, can be combined with other methods that simulate transport for macro structures and is visually equivalent to SSAO in the worst case without introducing new artifacts. Since our method works in screen space, it does not depend on the geometric complexity. Plausible directional occlusion and indirect lighting effects can be displayed for large and fully dynamic scenes at real-time frame rates.

Abstract Glare is a consequence of light scattered within the human eye when looking at bright light sources. This effect can be exploited for tone mapping since adding glare to the depiction of high‐dynamic range (HDR) imagery on a low‐dynamic range (LDR) medium can dramatically increase perceived contrast. Even though most, if not all, subjects report perceiving glare as a bright pattern that fluctuates in time, up to now it has only been modeled as a static phenomenon. We argue that the temporal properties of glare are a strong means to increase perceived brightness and to produce realistic and attractive renderings of bright light sources. Based on the anatomy of the human eye, we propose a model that enables real‐time simulation of dynamic glare on a GPU. This allows an improved depiction of HDR images on LDR media for interactive applications like games, feature films, or even by adding movement to initially static HDR images. By conducting psychophysical studies, we validate that our method improves perceived brightness and that dynamic glare‐renderings are often perceived as more attractive depending on the chosen scene.

In this article we evaluate the use of approximate visibility for efficient global illumination. Traditionally, accurate visibility is used in light transport. However, the indirect illumination we perceive on a daily basis is rarely of high-frequency nature, as the most significant aspect of light transport in real-world scenes is diffuse, and thus displays a smooth gradation. This raises the question of whether accurate visibility is perceptually necessary in this case. To answer this question, we conduct a psychophysical study on the perceptual influence of approximate visibility on indirect illumination. This study reveals that accurate visibility is not required and that certain approximations may be introduced.

Recent approaches to global illumination for dynamic scenes achieve interactive frame rates by using coarse approximations to geometry, lighting, or both, which limits scene complexity and rendering quality. High-quality global illumination renderings of complex scenes are still limited to methods based on ray tracing. While conceptually simple, these techniques are computationally expensive. We present an efficient and scalable method to compute global illumination solutions at interactive rates for complex and dynamic scenes. Our method is based on parallel final gathering running entirely on the GPU. At each final gathering location we perform micro-rendering: we traverse and rasterize a hierarchical point-based scene representation into an importance-warped micro-buffer , which allows for BRDF importance sampling. The final reflected radiance is computed at each gathering location using the micro-buffers and is then stored in image-space. We can trade quality for speed by reducing the sampling rate of the gathering locations in conjunction with bilateral upsampling. We demonstrate the applicability of our method to interactive global illumination, the simulation of multiple indirect bounces, and to final gathering from photon maps.

Effective digital content creation tools must be both efficient in the interactions they provide but also allow full user control. There may be occasions, when art direction requires changes that contradict physical laws. In particular, it is known that physical correctness of reflections for the human observer is hard to assess. For many centuries, traditional artists have exploited this fact to depict reflections that lie outside the realm of physical possibility. However, a system that gives explicit control of this effect to digital artists has not yet been described. This paper introduces a system that transforms physically correct reflections into art-directed reflections, as specified by reflection constraints . The system introduces a taxonomy of reflection editing operations, using an intuitive user interface, that works directly on the reflecting surfaces with real-time visual feedback using a GPU. A user study shows how such a system can allow users to quickly manipulate reflections according to an art direction task.

We present a method for interactive computation of indirect illumination in large and fully dynamic scenes based on approximate visibility queries. While the high-frequency nature of direct lighting requires accurate visibility, indirect illumination mostly consists of smooth gradations, which tend to mask errors due to incorrect visibility. We exploit this by approximating visibility for indirect illumination with imperfect shadow maps ---low-resolution shadow maps rendered from a crude point-based representation of the scene. These are used in conjunction with a global illumination algorithm based on virtual point lights enabling indirect illumination of dynamic scenes at real-time frame rates. We demonstrate that imperfect shadow maps are a valid approximation to visibility, which makes the simulation of global illumination an order of magnitude faster than using accurate visibility.

We present a new approach for enhancing local scene contrast by unsharp masking over arbitrary surfaces under any form of illumination. Our adaptation of a well-known 2D technique to 3D interactive scenarios is designed to aid viewers in tasks like understanding complex or detailed geometric models, medical visualization and navigation in virtual environments. Our holistic approach enhances the depiction of various visual cues, including gradients from surface shading, surface reflectance, shadows, and highlights, to ease estimation of viewpoint, lighting conditions, shapes of objects and their world-space organization. Motivated by recent perceptual findings on 3D aspects of the Cornsweet illusion, we create scene coherent enhancements by treating cues in terms of their 3D context; doing so has a stronger effect than approaches that operate in a 2D image context and also achieves temporal coherence. We validate our unsharp masking in 3D with psychophysical experiments showing that the enhanced images are perceived to have better contrast and are preferred over unenhanced originals. Our operator runs at real-time rates on a GPU and the effect is easily controlled interactively within the rendering pipeline.