arXiv subjects

Dieter Schmalstieg

Publications and source records attributed to Dieter Schmalstieg.

At least 19 recordsLinked to original sources

arXiv2026

SnapPhysics: A Physics-Aware Scene Graph from a Single View for Interactive Mixed Reality Scenes

We propose SnapPhysics, a training-free framework that reconstructs 3D objects and estimates their physical properties such as mass, friction, and center of gravity from a single image. For physically coherent interactions in mixed reality (MR), such properties are as important as geometry. Prior approaches infer them by analyzing object dynamics in video, which is computationally costly, or by querying vision-language models (VLMs) on single images, which lacks geometric grounding and inter-object relationships. We address these limitations by combining instance-level 3D reconstruction and spatial alignment with a physics-aware scene graph that encodes these relationships and per-object metric geometry as structured context for VLM-based property reasoning. Experiments on 3D-FRONT show that SnapPhysics improves scene-level F-Score by 18.6% over the best learning-based method, and on real captured scenes with ground-truth mass, it reduces the mean absolute log difference error (mALDE) by up to 20.5% and improves log-scale correlation ($r^2_{\mathrm{ls}}$) by up to 19.6% over VLM-only estimation. SnapPhysics enables physically interactive MR experiences without manual parameter tuning. Project page: https://snapphysics-ismar2026.github.io/.

cs.CV↗

arXiv2026

PhGS: Post-Hoc Pruning and Refinement of Single-View Feed-Forward 3D Gaussian Reconstructions

Recent single-view feed-forward 3D Gaussian Splatting (3DGS) generation predicts a fixed number of Gaussians per camera ray, introducing severe spatial redundancy. Most existing compaction strategies target multi-view setups to exploit cross-view consistency and are incompatible with single-image models. Instead of retraining the base feed-forward network to directly output compact representations, our insight is to keep the base models frozen and apply post-hoc pruning and recurrent refinement to the generated Gaussians. Consequently, we propose a backbone-agnostic compaction pipeline for single-view feed-forward 3DGS that couples an importance-score-based pruning mechanism with a trainable, lightweight recurrent refinement module, which iteratively updates the surviving primitives to restore image quality. Our results demonstrate seamless integration with existing baselines while preserving novel-view rendering fidelity and achieving high memory reduction. Furthermore, our method supports flexible inference-time keep ratios for application needs.

cs.CV↗

arXiv2026

Amulet: Frame Extrapolation Through Sparse Layered Scene Representation and Adaptive Shading

We introduce Amulet, a rendering method that transforms a scene into a sparse, tiled and layered intermediate scene representation (cache) for high-frequency frame extrapolation. In contrast to reprojection-based techniques, Amulet explicitly rasterizes and stores potentially visible geometry in its layered image-space cache, allowing accurate shading and inpainting of newly disoccluded regions without hallucination. Our key contribution is a cache that is predictively filled with shading information for future views, amortized over multiple current frames. Novel views are synthesized by hierarchically traversing the cache front to back and refining stale or missing shading on the fly. Using a predictive, gradient-based scheduler that assigns lifetimes for each tile, we enable adaptive shading updates under motion and dynamic lighting. Amulet decouples the rasterization and shading rate from the refresh rate of the display. In many scenarios, our cache can use a single shaded frame to synthesize multiple extrapolated frames with only a few localized updates. In a typical application, we extrapolate a 60 Hz shading rate to a 240 Hz display. Amulet achieves up to 250 Hz at 4K resolution and is competitive with state-of-the-art frame generation methods, including DLSS and neural-flow approaches, in multiple metrics. Amulet explores the design space of sparse layered image-space representation. It enables accurate, non-neural multi frame extrapolation with explicit handling of disocclusions. Our findings show that Amulet can extrapolate many more frames than contemporary methods with high quality, rivaling latency-bound frame interpolation methods with similar quality in many scenes.

cs.GR↗

arXiv2026

3D-GIMP: When 3D Gaussian Inpainting Meets PatchMatch

Recent advances in 3D scene editing have leveraged iterative diffusion models to update input views. However, this process is computationally expensive and struggles to produce sharp details. Meanwhile, ``hallucination drift'' frequently introduces multi-view inconsistencies, leading to structural artifacts when rendering novel viewpoints. To address this problem, we present 3D-GIMP (3D Gaussian Inpainting Meets Patch Matching), a novel hybrid paradigm designed for high-fidelity object removal in 3D Gaussian Splatting. Instead of diffusing every view, 3D-GIMP performs a single generative inpainting on a key reference view, which serves as an appearance prior. We then introduce a 3D-aware PatchMatch algorithm to propagate these reference textures across all remaining views via correspondence matching, effectively bypassing the stochastic nature of frame-by-frame diffusion. By prioritizing reconstructive consistency over iterative generation, 3D-GIMP maintains high-frequency details across arbitrary resolutions while ensuring a mathematically consistent 3D reconstruction. Our experiments demonstrate that 3D-GIMP not only achieves competitive inpainting quality as previous methods using diffusion in multiple views, but also outperforms these methods in rendering speed and view consistency.

cs.CV↗

arXiv2026

Think While You Map: Asynchronous Vision-Language Agents for Incremental 3D Scene Graphs

Open-vocabulary 3D scene graph methods typically operate in two stages: first reconstruct, then enrich with vision-language models, leaving the graph unqueryable during exploration. We argue that this sequential coupling is unnecessary and propose an asynchronous architecture in which lightweight online mapping runs concurrently with heavyweight semantic refinement. A probabilistic voxel-based backbone maintains stable object identities incrementally, while background VLM agents progressively enrich the graph. This framework resolves duplicate object tracks through semantic loop closure, attaches fine-grained visual attributes and derives spatial relations between objects. A multi-target frame scheduler amortizes VLM cost by selecting a small set of informative frames that jointly cover multiple targets. The resulting scene graph is queryable during exploration and grows in semantic richness over time. Our method matches or outperforms existing open-vocabulary 3D scene graph methods on semantic segmentation (ScanNet, Replica) and surpasses the prior state-of-the-art across three visual grounding benchmarks (Sr3D+, Nr3D, ScanRefer) by 15.3 to 18.8 A@0.25. Project page: https://denizbickici.github.io/thinkgraphs/

cs.CV↗

arXiv2026

Ambient Analytics: Calm Technology for Immersive Visualization and Sensemaking

Augmented reality has great potential for embedding data visualizations in the world around the user. While this can enhance users' understanding of their surroundings, it also bears the risk of overwhelming their senses with a barrage of information. In contrast, calm technologies aim to place information in the user's attentional periphery, minimizing cognitive load instead of demanding focused engagement. In this column, we explore how visualizations can be harmoniously integrated into our everyday life through augmented reality, progressing from visual analytics to ambient analytics.

cs.HC↗

arXiv2026

A LoD of Gaussians: Unified Training and Rendering for Ultra-Large Scale Reconstruction with External Memory

Gaussian Splatting has emerged as a high-performance technique for novel view synthesis, enabling real-time rendering and high-quality reconstruction of small scenes. However, scaling to larger environments has so far relied on partitioning the scene into chunks -- a strategy that introduces artifacts at chunk boundaries, complicates training across varying scales, and is poorly suited to unstructured scenarios such as city-scale flyovers combined with street-level views. Moreover, rendering remains fundamentally limited by GPU memory, as all visible chunks must reside in VRAM simultaneously. We introduce A LoD of Gaussians, a framework for training and rendering ultra-large-scale Gaussian scenes on a single consumer-grade GPU -- without partitioning. Our method stores the full scene out-of-core (e.g., in CPU memory) and trains a Level-of-Detail (LoD) representation directly, dynamically streaming only the relevant Gaussians. A hybrid data structure combining Gaussian hierarchies with Sequential Point Trees enables efficient, view-dependent LoD selection, while a lightweight caching and view scheduling system exploits temporal coherence to support real-time streaming and rendering. Together, these innovations enable seamless multi-scale reconstruction and interactive visualization of complex scenes -- from broad aerial views to fine-grained ground-level details.

cs.GR↗

arXiv2026

Situated Brushing and Linking in Virtual and Augmented Reality

In traditional visual analysis, brushing and linking is commonly used to visually connect multiple views using highlighting techniques. However, brushing and linking has rarely been used in situated analytics, which uses visualizations to analyze data in the context of physical referents. In situated analytics, data representations must be visually linked to real-world objects. Previous work has assessed situated brushing and linking in a virtual reality simulation of a supermarket scenario. Here, we replicate and extend the previous approach by studying brushing and linking in an actual physical space with augmented reality, while further improving the highlighting techniques. Using a video see-through display, we compare augmented reality with virtual reality. Results suggest that AR performs better in time and accuracy, but the effectiveness of the techniques varies by condition. These results provide a new framing of how the real-world stimuli matter in situated analytics.

cs.HC↗

arXiv2026

Hybrid User Interfaces: Past, Present, and Future of Complementary Cross-Device Interaction in Mixed Reality

We investigate hybrid user interfaces (HUIs), aiming to establish a cohesive understanding and adopt consistent terminology for this nascent research area. HUIs combine heterogeneous devices in complementary roles, leveraging the distinct benefits of each. Our work focuses on cross-device interaction between 2D devices and mixed reality environments, which are particularly compelling, leveraging the familiarity of traditional 2D platforms while providing spatial awareness and immersion. Although such HUIs have been prominently explored in the context of mixed reality by prior work, we still lack a cohesive understanding of the unique design possibilities and challenges of such combinations, resulting in a fragmented research landscape. We conducted a systematic survey and present a taxonomy of HUIs that combine conventional display technology and mixed reality environments. Based on this, we discuss past and current challenges, the evolution of definitions, and prospective opportunities to tie together the past 30 years of research with our vision of future HUIs.

cs.HC↗

arXiv2025

SOF: Sorted Opacity Fields for Fast Unbounded Surface Reconstruction

Recent advances in 3D Gaussian representations have significantly improved the quality and efficiency of image-based scene reconstruction. Their explicit nature facilitates real-time rendering and fast optimization, yet extracting accurate surfaces - particularly in large-scale, unbounded environments - remains a difficult task. Many existing methods rely on approximate depth estimates and global sorting heuristics, which can introduce artifacts and limit the fidelity of the reconstructed mesh. In this paper, we present Sorted Opacity Fields (SOF), a method designed to recover detailed surfaces from 3D Gaussians with both speed and precision. Our approach improves upon prior work by introducing hierarchical resorting and a robust formulation of Gaussian depth, which better aligns with the level-set. To enhance mesh quality, we incorporate a level-set regularizer operating on the opacity field and introduce losses that encourage geometrically-consistent primitive shapes. In addition, we develop a parallelized Marching Tetrahedra algorithm tailored to our opacity formulation, reducing meshing time by up to an order of magnitude. As demonstrated by our quantitative evaluation, SOF achieves higher reconstruction accuracy while cutting total processing time by more than a factor of three. These results mark a step forward in turning efficient Gaussian-based rendering into equally efficient geometry extraction.

cs.GR↗

arXiv2025

NeuralPVS: Learned Estimation of Potentially Visible Sets

Real-time visibility determination in expansive or dynamically changing environments has long posed a significant challenge in computer graphics. Existing techniques are computationally expensive and often applied as a precomputation step on a static scene. We present NeuralPVS, the first deep-learning approach for visibility computation that efficiently determines from-region visibility in a large scene, running at approximately 100 Hz processing with less than $1\%$ missing geometry. This approach is possible by using a neural network operating on a voxelized representation of the scene. The network's performance is achieved by combining sparse convolution with a 3D volume-preserving interleaving for data compression. Moreover, we introduce a novel repulsive visibility loss that can effectively guide the network to converge to the correct data distribution. This loss provides enhanced robustness and generalization to unseen scenes. Our results demonstrate that NeuralPVS outperforms existing methods in terms of both accuracy and efficiency, making it a promising solution for real-time visibility computation.

cs.GR↗

arXiv2025

3D Gaussian Modeling and Ray Marching of OpenVDB datasets for Scientific Visualization

3D Gaussians are currently being heavily investigated for their scene modeling and compression abilities. In 3D volumes, their use is being explored for representing dense volumes as sparsely as possible. However, most of these methods begin with a memory inefficient data format. Specially in Scientific Visualization(SciVis), where most popular formats are dense-grid data structures that store every grid cell, irrespective of its contribution. OpenVDB library and data format were introduced for representing sparse volumetric data specifically for visual effects use cases such as clouds, fire, fluids etc. It avoids storing empty cells by masking them during storage. It presents an opportunity for use in SciVis, specifically as a modeling framework for conversion to 3D Gaussian particles for further compression and for a unified modeling approach for different scientific volume types. This compression head-start is non-trivial and this paper would like to present this with a rendering algorithm based on line integration implemented in OptiX8.1 for calculating 3D Gaussians contribution along a ray for optical-depth accumulation. For comparing the rendering results of our ray marching Gaussians renderer, we also implement a SciVis style primary-ray only NanoVDB HDDA based ray marcher for OpenVDB voxel grids. Finally, this paper also explores application of this Gaussian model to formats of volumes other than regular grids, such as AMR volumes and point clouds, using internal representation of OpenVDB grid class types for data hierarchy and subdivision structure.

cs.GR↗

arXiv2025

Long-Term Experiences From Working with Extended Reality in the Wild

Extended Reality (XR) is increasingly used as a productivity tool and recent commercial XR devices have even been specifically designed as productivity tools, or, at least, are heavily advertised for such purposes, such as the Apple Vision Pro (AVP), which has now been available for more than one year. In spite of what marketing suggests, research still lacks an understanding of the long-term usage of such devices in ecologically valid everyday settings, as most studies are conducted in very controlled environments. Therefore, we conducted interviews with ten AVP users to better understand how experienced users engage with the device, and which limitations persist. Our participants report that XR can increase productivity and that they got used to the device after some time. Yet, a range of limitations persist that might hinder the widespread use of XR as a productivity tool, such as a lack of native applications, difficulties when integrating XR into current workflows, and limited possibilities to adapt and customize the XR experience.

cs.HC↗

arXiv2025

IntelliCap: Intelligent Guidance for Consistent View Sampling

Novel view synthesis from images, for example, with 3D Gaussian splatting, has made great progress. Rendering fidelity and speed are now ready even for demanding virtual reality applications. However, the problem of assisting humans in collecting the input images for these rendering algorithms has received much less attention. High-quality view synthesis requires uniform and dense view sampling. Unfortunately, these requirements are not easily addressed by human camera operators, who are in a hurry, impatient, or lack understanding of the scene structure and the photographic process. Existing approaches to guide humans during image acquisition concentrate on single objects or neglect view-dependent material characteristics. We propose a novel situated visualization technique for scanning at multiple scales. During the scanning of a scene, our method identifies important objects that need extended image coverage to properly represent view-dependent appearance. To this end, we leverage semantic segmentation and category identification, ranked by a vision-language model. Spherical proxies are generated around highly ranked objects to guide the user during scanning. Our results show superior performance in real scenes compared to conventional view sampling strategies.

cs.CV↗

arXiv2025

AAA-Gaussians: Anti-Aliased and Artifact-Free 3D Gaussian Rendering

Although 3D Gaussian Splatting (3DGS) has revolutionized 3D reconstruction, it still faces challenges such as aliasing, projection artifacts, and view inconsistencies, primarily due to the simplification of treating splats as 2D entities. We argue that incorporating full 3D evaluation of Gaussians throughout the 3DGS pipeline can effectively address these issues while preserving rasterization efficiency. Specifically, we introduce an adaptive 3D smoothing filter to mitigate aliasing and present a stable view-space bounding method that eliminates popping artifacts when Gaussians extend beyond the view frustum. Furthermore, we promote tile-based culling to 3D with screen-space planes, accelerating rendering and reducing sorting costs for hierarchical rasterization. Our method achieves state-of-the-art quality on in-distribution evaluation sets and significantly outperforms other approaches for out-of-distribution views. Our qualitative evaluations further demonstrate the effective removal of aliasing, distortions, and popping artifacts, ensuring real-time, artifact-free rendering.

cs.GR↗

arXiv2025

3D Gaussian Particle Approximation of VDB Datasets: A Study for Scientific Visualization

The complexity and scale of Volumetric and Simulation datasets for Scientific Visualization(SciVis) continue to grow. And the approaches and advantages of memory-efficient data formats and storage techniques for such datasets vary. OpenVDB library and its VDB data format excels in memory efficiency through its hierarchical and dynamic tree structure, with active and inactive sub-trees for data storage. It is heavily used in current production renderers for both animation and rendering stages in VFX pipelines and photorealistic rendering of volumes and fluids. However, it still remains to be fully leveraged in SciVis where domains dealing with sparse scalar fields like porous media, time varying volumes such as tornado and weather simulation or high resolution simulation of Computational Fluid Dynamics present ample number of large challenging data sets. The goal of this paper hence is not only to explore the use of OpenVDB in SciVis but also to explore a level of detail(LOD) technique using 3D Gaussian particles approximating voxel regions. For rendering, we utilize NVIDIA OptiX library for ray marching through the Gaussians particles. Data modeling using 3D Gaussians has been very popular lately due to success in stereoscopic image to 3D scene conversion using Gaussian Splatting and Gaussian approximation and mixture models aren't entirely new in SciVis as well. Our work explores the integration with rendering software libraries like OpenVDB and OptiX to take advantage of their built-in memory compaction and hardware acceleration features, while also leveraging the performance capabilities of modern GPUs. Thus, we present a SciVis rendering approach that uses 3D Gaussians at varying LOD in a lossy scheme derived from VDB datasets, rather than focusing on photorealistic volume rendering.

cs.GR↗

arXiv2024

Envisioning Situated Visualizations of Environmental Footprints in an Urban Environment

We present the results of a brainstorming exercise focused on how situated visualizations could be used to better understand the state of the environment and our personal behavioral impact on it. Specifically, we conducted a day long workshop in the French city of Bordeaux where we envisioned situated visualizations of urban environmental footprints. We explored the city and took photos and notes about possible situated visualizations of environmental footprints that could be embedded near places, people, or objects of interest. We found that our designs targeted four purposes and used four different methods that could be further explored to test situated visualizations for the protection of the environment.

cs.HC↗

arXiv2024

Visual Highlighting for Situated Brushing and Linking

Brushing and linking is widely used for visual analytics in desktop environments. However, using this approach to link many data items between situated (e.g., a virtual screen with data) and embedded views (e.g., highlighted objects in the physical environment) is largely unexplored. To this end, we study the effectiveness of visual highlighting techniques in helping users identify and link physical referents to brushed data marks in a situated scatterplot. In an exploratory virtual reality user study (N=20), we evaluated four highlighting techniques under different physical layouts and tasks. We discuss the effectiveness of these techniques, as well as implications for the design of brushing and linking operations in situated analytics.

cs.HC↗