We present a Directional Consistency (DC)-driven Adaptive Density Control (ADC) for 3D Gaussian Splatting (DC4GS). Whereas the conventional ADC bases its primitive splitting on the magnitudes of positional gradients, we further incorporate the DC of the gradients into ADC, and realize it through the angular coherence of the gradients. Our DC better captures local structural complexities in ADC, avoiding redundant splitting. When splitting is required, we again utilize the DC to define optimal split positions so that sub-primitives best align with the local structures than the conventional random placement. As a consequence, our DC4GS greatly reduces the number of primitives (up to 30\% in our experiments) than the existing ADC, and also enhances reconstruction fidelity greatly.

This paper presents the model and rendering algorithm of Potentially Visible Hidden Volumes (PVHVs) for multi-view image warping. PVHVs are 3D volumes that are occluded at a known source view, but potentially visible at novel views. Given a bound of novel views, we define PVHVs using the edges of foreground fragments from the known view and the bound of novel views. PVHVs can be used to batch-test the visibilities of source fragments without iterating individual novel views in multi-fragment rendering, and thereby, cull redundant fragments prior to warping. We realize the model of PVHVs in Depth Peeling (DP). Our Effective Depth Peeling (EDP) can reduce the number of completely hidden fragments, capture important fragments early, and reduce warping cost. We demonstrate the benefit of our PVHVs and EDP in terms of memory, quality, and performance in multi-view warping.

This article presents a real-time bokeh rendering technique that splats pre-computed sprites but takes dynamic visibilities and intrinsic appearances into account at runtime. To attain alias-free looks without excessive sampling on a lens, the visibilities of strong highlights are densely sampled using rasterization, while regular objects are sparsely sampled using conventional defocus-blur rendering. The intrinsic appearance is dynamically transformed from a precomputed look-up table, which encodes radial aberrations against image distances in a compact 2D texture. Our solution can render complex bokeh effects without undersampling artifacts in real time, and greatly improve the photorealism of defocus-blur rendering.

Abstract This paper presents a scalable online occlusion culling algorithm, which significantly improves the previous raster occlusion culling using object‐level bounding volume hierarchy. Given occluders found with temporal coherence, we find and rasterize coarse groups of potential occludees in the hierarchy. Within the rasterized bounds, per‐pixel ray casting tests fine‐grained visibilities of every individual occludees. We further propose acceleration techniques including the read‐back of counters for tightly‐packed multidrawing and occluder filtering. Our solution requires only constant draw calls for batch occlusion tests, while avoiding costly iteration for hierarchy traversal. Our experiments prove our solution outperforms the existing solutions in terms of scalability, culling efficiency, and occlusion‐query performance.

In this paper, we propose the first end-to-end convolutional neural network (CNN) architecture, Defocus Map Estimation Network (DMENet), for spatially varying defocus map estimation. To train the network, we produce a novel depth-of-field (DOF) dataset, SYNDOF, where each image is synthetically blurred with a ground-truth depth map. Due to the synthetic nature of SYNDOF, the feature characteristics of images in SYNDOF can differ from those of real defocused photos. To address this gap, we use domain adaptation that transfers the features of real defocused photos into those of synthetically blurred ones. Our DMENet consists of four subnetworks: blur estimation, domain adaptation, content preservation, and sharpness calibration networks. The subnetworks are connected to each other and jointly trained with their corresponding supervisions in an end-to-end manner. Our method is evaluated on publicly available blur detection and blur estimation datasets and the results show the state-of-the-art performance.In this paper, we propose the first end-to-end convolutional neural network (CNN) architecture, Defocus Map Estimation Network (DMENet), for spatially varying defocus map estimation. To train the network, we produce a novel depth-of-field (DOF) dataset, SYNDOF, where each image is synthetically blurred with a ground-truth depth map. Due to the synthetic nature of SYNDOF, the feature characteristics of images in SYNDOF can differ from those of real defocused photos. To address this gap, we use domain adaptation that transfers the features of real defocused photos into those of synthetically blurred ones. Our DMENet consists of four subnetworks: blur estimation, domain adaptation, content preservation, and sharpness calibration networks. The subnetworks are connected to each other and jointly trained with their corresponding supervisions in an end-to-end manner. Our method is evaluated on publicly available blur detection and blur estimation datasets and the results show the state-of-the-art performance.

Abstract We present a scalable solution to render complex scenes from a large amount of viewpoints. While previous approaches rely either on a scene or a view hierarchy to process multiple elements together, we make full use of both, enabling sublinear performance in terms of views and scene complexity. By concurrently traversing the hierarchies, we efficiently find shared information among views to amortize rendering costs. One example application is many‐light global illumination. Our solution accelerates shadow map generation for virtual point lights, whose number can now be raised to over a million while maintaining interactive rates.

Abstract Far‐field diffraction can be evaluated using the Discrete Fourier Transform (DFT) in image space but it is costly due to its dense sampling. We propose a technique based on a closed‐form solution of the continuous Fourier transform for simple vector primitives (quads) and propose a hierarchical and progressive evaluation to achieve real‐time performance. Our method is able to simulate diffraction effects in optical systems and can handle varying visibility due to dynamic light sources. Furthermore, it seamlessly extends to near‐field diffraction. We show the benefit of our solution in various applications, including realistic real‐time glare and bloom rendering.

Abstract This paper presents a novel heterogeneous volume deformation technique and an intuitive volume animation authoring framework. Our volume deformation extends the previous technique based on moving least squares with a density‐aware weighting metric for data‐driven importance control and efficient upsampling‐based volume synthesis. For user interaction, we present an intuitive visual metaphor and interaction schemes to support effective spatiotemporal editing of volume deformation animation. Our framework is implemented fully on graphics processors and thus suitable for quick‐and‐easy prototyping of volume deformation with improved controllability.

This article presents an iterative backward-warping technique and its applications. It predictively synthesizes depth buffers for novel views. Our solution is based on a fixed-point iteration that converges quickly in practice. Unlike the previous techniques, our solution is a pure backward warping without using bidirectional sources. To efficiently seed the iterative process, we also propose a tight bounding method for motion vectors. Non-convergent depth holes are inpainted via deep depth buffers. Our solution works well with arbitrarily distributed motion vectors under moderate motions. Many scenarios can benefit from our depth warping. As an application, we propose a highly scalable image-based occlusion-culling technique, achieving a significant speed-up compared to the state of the art. We also demonstrate the benefit of our solution in multi-view soft-shadow generation.

We present an efficient ray-tracing technique to render bokeh effects produced by parametric aspheric lenses. Contrary to conventional spherical lenses, aspheric lenses do generally not permit a simple closed-form solution of ray-surface intersections. We propose a numerical root-finding approach, which uses tight proxy surfaces to ensure a good initialization and convergence behavior. Additionally, we simulate mechanical imperfections resulting from the lens fabrication via a texture-based approach. Fractional Fourier transform and spectral dispersion add additional realism to the synthesized bokeh effect. Our approach is well-suited for execution on graphics processing units (GPUs) and we demonstrate complex defocus-blur and lens-flare effects.

In this work, we present a technique based on image saliency analysis to improve the conspicuity of the foreground augmentation to the background real world medium by adjusting the local brightness contrast. The proposed technique is implemented on a mobile platform considering the usage nature of AR. The saliency computation is carried out for the augmented object's representative color rather than all the pixels, and searching and adjusting over only a discrete number of brightness levels to produce the highest contrast saliency, thereby making near-real time computation possible. Thus, while the resulting imagery may not be optimal due to such a simplification, our tests showed that the visibility was still significantly improved without much difference to the "optimal" ground truth in terms of correctly perceiving and recognizing the augmented information

Abstract This paper presents a GPU‐based rendering algorithm for real‐time defocus blur effects, which significantly improves the accumulation buffering. The algorithm combines three distinctive techniques: (1) adaptive discrete geometric level of detail (LOD), made popping‐free by blending visibility samples across the two adjacent geometric levels; (2) adaptive visibility/shading sampling via sample reuse; (3) visibility supersampling via height‐field ray casting. All the three techniques are seamlessly integrated to lower the rendering cost of smooth defocus blur with high visibility sampling rates, while maintaining most of the quality of brute‐force accumulation buffering.

Abstract We present a practical real‐time approach for rendering lens‐flare effects. While previous work employed costly ray tracing or complex polynomial expressions, we present a coarser, but also significantly faster solution. Our method is based on a first‐order approximation of the ray transfer in an optical system, which allows us to derive a matrix that maps lens flare‐producing light rays directly to the sensor. The resulting approach is easy to implement and produces physically‐plausible images at high framerates on standard off‐the‐shelf graphics hardware.

Visualization techniques often use color to present categorical differences to a user. When selecting a color palette, the perceptual qualities of color need careful consideration. Large coherent groups visually suppress smaller groups and are often visually dominant in images. This paper introduces the concept of class visibility used to quantitatively measure the utility of a color palette to present coherent categorical structure to the user. We present a color optimization algorithm based on our class visibility metric to make categorical differences clearly visible to the user. We performed two user experiments on user preference and visual search to validate our visibility measure over a range of color palettes. The results indicate that visibility is a robust measure, and our color optimization can increase the effectiveness of categorical data visualizations.

This article evaluates the usability of motion sensing-based interaction on a mobile platform using image browsing as a representative task. Three types of interfaces, a physical button interface, a motion-sensing interface using a high-precision commercial 3D motion tracker, and a motion-sensing interface using an in-house low-cost 3D motion tracker, are compared in terms of task performance and subjective preference. Participants were provided with prolonged training over 20 days, in order to compensate for the participants’ unfamiliarity with the motion-sensing interfaces. Experimental results showed that the participants’ task performance and subjective preference for the two motion-sensing interfaces were initially low, but they rapidly improved with training and soon approached the level of the button interface. Furthermore, a recall test, which was conducted 4 weeks later, demonstrated that the usability gains were well retained in spite of the long time gap between uses. Overall, these findings highlight the potential of motion-based interaction as an intuitive interface for mobile devices.

Lens flare is caused by light passing through a photographic lens system in an unintended way. Often considered a degrading artifact, it has become a crucial component for realistic imagery and an artistic means that can even lead to an increased perceived brightness. So far, only costly offline processes allowed for convincing simulations of the complex light interactions. In this paper, we present a novel method to interactively compute physically-plausible flare renderings for photographic lenses. The underlying model covers many components that are important for realism, such as imperfections, chromatic and geometric lens aberrations, and antireflective lens coatings. Various acceleration strategies allow for a performance/quality tradeoff, making our technique applicable both in real-time applications and in high-quality production rendering. We further outline artistic extensions to our system.

We present a novel rendering system for defocus blur and lens effects. It supports physically-based rendering and outperforms previous approaches by involving a novel GPU-based tracing method. Our solution achieves more precision than competing real-time solutions and our results are mostly indistinguishable from offline rendering. Our method is also more general and can integrate advanced simulations, such as simple geometric lens models enabling various lens aberration effects. These latter is crucial for realism, but are often employed in artistic contexts, too. We show that available artistic lenses can be simulated by our method. In this spirit, our work introduces an intuitive control over depth-of-field effects. The physical basis is crucial as a starting point to enable new artistic renderings based on a generalized focal surface to emphasize particular elements in the scene while retaining a realistic look. Our real-time solution provides realistic, as well as plausible expressive results.

We present a GPU-based real-time rendering method that simulates high-quality depth-of-field effects, similar in quality to multiview accumulation methods. Most real-time approaches have difficulties to obtain good approximations of visibility and view-dependent shading due to the use of a single view image. Our method also avoids the multiple rendering of a scene, but can approximate different views by relying on a layered image-based scene representation. We present several performance and quality improvements, such as early culling, approximate cone tracing, and jittered sampling. Our method achieves artifact-free results for complex scenes and reasonable depth-of-field blur in real time.

This paper presents a real-time framework for computationally tracking objects visually attended by the user while navigating in interactive virtual environments. In addition to the conventional bottom-up (stimulus-driven) saliency map, the proposed framework uses top-down (goal-directed) contexts inferred from the user's spatial and temporal behaviors, and identifies the most plausibly attended objects among candidates in the object saliency map. The computational framework was implemented using GPU, exhibiting high computational performance adequate for interactive virtual environments. A user experiment was also conducted to evaluate the prediction accuracy of the tracking framework by comparing objects regarded as visually attended by the framework to actual human gaze collected with an eye tracker. The results indicated that the accuracy was in the level well supported by the theory of human cognition for visually identifying single and multiple attentive targets, especially owing to the addition of top-down contextual information. Finally, we demonstrate how the visual attention tracking framework can be applied to managing the level of details in virtual environments, without any hardware for head or eye tracking.

This article presents a real-time GPU-based post-filtering method for rendering acceptable depth-of-field effects suited for virtual reality. Blurring is achieved by nonlinearly interpolating mipmap images generated from a pinhole image. Major artifacts common in the post-filtering techniques such as bilinear magnification artifact, intensity leakage, and blurring discontinuity are practically eliminated via magnification with a circular filter, anisotropic mipmapping, and smoothing of blurring degrees. The whole framework is accelerated using GPU programs for constant and scalable real-time performance required for virtual reality. We also compare our method to recent GPU-based methods in terms of image quality and rendering performance.

Abstract We present a real‐time method for rendering a depth‐of‐field effect based on the per‐pixel layered splatting where source pixels are scattered on one of the three layers of a destination pixel. In addition, the missing information behind foreground objects is filled with an additional image of the areas occluded by nearer objects. The method creates high‐quality depth‐of‐field results even in the presence of partial occlusion, without major artifacts often present in the previous real‐time methods. The method can also be applied to simulating defocused highlights. The entire framework is accelerated by GPU, enabling real‐time post‐processing for both off‐line and interactive applications.

This article reports two human experiments to investigate the effects of visual cues and sustained attention on spatial presence over a period of prolonged exposure in virtual environments. Inspired by the two functional subsystems subserving spatial and object vision in the human brain, visual cues and sustained attention were each classified into spatial and object cues, and spatial and non-spatial attention, respectively. In the first experiment, the effects of visual cues on spatial presence were examined when subjects were exposed to virtual environments configured with combinations of spatial and object cues. It was found that both types of visual cues enhanced spatial presence with saturation over a period of prolonged exposure, but the contribution of spatial cues became more relevant with longer exposure time. In the second experiment, subjects were asked to carry out two tasks involving sustained spatial attention and sustained non-spatial attention. We observed that spatially directed attention improved spatial presence more than non-spatially directed attention did. Furthermore, spatial attention had a positive interaction with detailed object cues.

This paper presents a real-time framework for computationally tracking objects visually attended by the user while navigating in interactive virtual environments. In addition to the conventional bottom-up (stimulus-driven) features, the framework also uses topdown (goal-directed) contexts to predict the human gaze. The framework first builds feature maps using preattentive features such as luminance, hue, depth, size, and motion. The feature maps are then integrated into a single saliency map using the center-surround difference operation. This pixel-level bottom-up saliency map is converted to an object-level saliency map using the item buffer. Finally, the top-down contexts are inferred from the user's spatial and temporal behaviors during interactive navigation and used to select the most plausibly attended object among candidates produced in the object saliency map. The computational framework was implemented using the GPU and exhibited extremely fast computing performance (5.68 msec for a 256X256 saliency map), substantiating its adequacy for interactive virtual environments. A user experiment was also conducted to evaluate the prediction accuracy of the visual attention tracking framework with respect to actual human gaze data. The attained accuracy level was well supported by the theory of human cognition for visually identifying a single and multiple attentive targets, especially due to the addition of top-down contextual information. The framework can be effectively used for perceptually based rendering without employing an expensive eye tracker, such as providing the depth-of-field effects and managing the level-of-detail in virtual environments.

Presence is one of the goals of many virtual reality systems. Historically, in the context of virtual reality, the concept of presence has been associated much with spatial perception (bottom up process) as its informal definition of "feeling of being there" suggests. However, recent studies in presence have challenged this view and attempted to widen the concept to include psychological immersion, thus linking more high level elements (processed in a top down fashion) to presence such as story and plots, flow, attention and focus, identification with the characters, emotion, etc. In this paper, we experimentally studied the relationship between two content elements, each representing the two axis of the presence dichotomy, perceptual cues for spatial presence and sustained attention for (psychological) immersion. Our belief was that spatial perception or presence and a top down processed concept such as voluntary attention have only a very weak relationship, thus our experimental hypothesis was that sustained attention would positively affect spatial presence in a virtual environment with impoverished perceptual cues, but have no effect in an environment rich in them. In order to confirm the existence of the sustained attention in the experiment, fMRI of the subjects were taken and analyzed as well. The experimental results showed that that attention had no effect on spatial presence, even in the environment with impoverished spatial cues.