arXiv subjects

Marc Stamminger

Publications and source records attributed to Marc Stamminger.

At least 19 recordsLinked to original sources

arXiv2026

Fus3D: Decoding Consolidated 3D Geometry from Feed-forward Geometry Transformer Latents

We propose a feed-forward method for dense Signed Distance Field (SDF) regression from unstructured image collections in less than three seconds, without camera calibration or post-hoc fusion. Our key insight is that the intermediate feature space of pretrained multi-view feed-forward geometry transformers already encodes a powerful joint world representation; yet, existing pipelines discard it, routing features through per-view prediction heads before assembling 3D geometry post-hoc, which discards valuable completeness information and accumulates inaccuracies. We instead perform 3D extraction directly from geometry transformer features via learned volumetric extraction: voxelized canonical embeddings that progressively absorb multi-view geometry information through interleaved cross- and self-attention into a structured volumetric latent grid. A simple convolutional decoder then maps this grid to a dense SDF. We additionally propose a scalable, validity-aware supervision scheme directly using SDFs derived from depth maps or 3D assets, tackling practical issues like non-watertight meshes. Our approach yields complete and well-defined distance values across sparse- and dense-view settings and demonstrates geometrically plausible completions. Code and further material can be found at https://lorafib.github.io/fus3d.

cs.CV↗

arXiv2025

SurfFill: Completion of LiDAR Point Clouds via Gaussian Surfel Splatting

LiDAR-captured point clouds are often considered the gold standard in active 3D reconstruction. While their accuracy is exceptional in flat regions, the capturing is susceptible to miss small geometric structures and may fail with dark, absorbent materials. Alternatively, capturing multiple photos of the scene and applying 3D photogrammetry can infer these details as they often represent feature-rich regions. However, the accuracy of LiDAR for featureless regions is rarely reached. Therefore, we suggest combining the strengths of LiDAR and camera-based capture by introducing SurfFill: a Gaussian surfel-based LiDAR completion scheme. We analyze LiDAR capturings and attribute LiDAR beam divergence as a main factor for artifacts, manifesting mostly at thin structures and edges. We use this insight to introduce an ambiguity heuristic for completed scans by evaluating the change in density in the point cloud. This allows us to identify points close to missed areas, which we can then use to grow additional points from to complete the scan. For this point growing, we constrain Gaussian surfel reconstruction to focus optimization and densification on these ambiguous areas. Finally, Gaussian primitives of the reconstruction in ambiguous areas are extracted and sampled for points to complete the point cloud. To address the challenges of large-scale reconstruction, we extend this pipeline with a divide-and-conquer scheme for building-sized point cloud completion. We evaluate on the task of LiDAR point cloud completion of synthetic and real-world scenes and find that our method outperforms previous reconstruction methods.

cs.CV↗

arXiv2025

MM-2FSK: Multimodal Frequency Shift Keying for Ultra-Efficient and Robust High-Resolution MIMO Radar Imaging

Accurate reconstruction of static and rapidly moving targets demands three-dimensional imaging solutions with high temporal and spatial resolution. Radar sensors are a promising sensing modality because of their fast capture rates and their independence from lighting conditions. To achieve high spatial resolution, MIMO radars with large apertures are required. Yet, they are infrequently used for dynamic scenarios due to significant limitations in signal processing algorithms. These limitations impose substantial hardware constraints due to their computational intensity and reliance on large signal bandwidths, ultimately restricting the sensor's capture rate. One solution of previous work is to use few frequencies only, which enables faster capture and requires less computation; however, this requires coarse knowledge of the target's position and works in a limited depth range only. To address these challenges, we extend previous work into the multimodal domain with MM-2FSK, which leverages an assistive optical depth sensing modality to obtain a depth prior, enabling high framerate capture with only few frequencies. We evaluate our method using various target objects with known ground truth geometry that is spatially registered to real millimeter-wave MIMO radar measurements. Our method demonstrates superior performance in terms of depth quality, being able to compete with the time- and resource-intensive measurements with many frequencies.

eess.SP↗

arXiv2025

Two-Stage Gaussian Splatting Optimization for Outdoor Scene Reconstruction

Outdoor scene reconstruction remains challenging due to the stark contrast between well-textured, nearby regions and distant backgrounds dominated by low detail, uneven illumination, and sky effects. We introduce a two-stage Gaussian Splatting framework that explicitly separates and optimizes these regions, yielding higher-fidelity novel view synthesis. In stage one, background primitives are initialized within a spherical shell and optimized using a loss that combines a background-only photometric term with two geometric regularizers: one constraining Gaussians to remain inside the shell, and another aligning them with local tangential planes. In stage two, foreground Gaussians are initialized from a Structure-from-Motion reconstruction, added and refined using the standard rendering loss, while the background set remains fixed but contributes to the final image formation. Experiments on diverse outdoor datasets show that our method reduces background artifacts and improves perceptual quality compared to state-of-the-art baselines. Moreover, the explicit background separation enables automatic, object-free environment map estimation, opening new possibilities for photorealistic outdoor rendering and mixed-reality applications.

cs.GR↗

arXiv2025

Towards Integrating Multi-Spectral Imaging with Gaussian Splatting

We present a study of how to integrate color (RGB) and multi-spectral imagery (red, green, red-edge, and near-infrared) into the 3D Gaussian Splatting (3DGS) framework, a state-of-the-art explicit radiance-field-based method for fast and high-fidelity 3D reconstruction from multi-view images. While 3DGS excels on RGB data, naive per-band optimization of additional spectra yields poor reconstructions due to inconsistently appearing geometry in the spectral domain. This problem is prominent, even though the actual geometry is the same, regardless of spectral modality. To investigate this, we evaluate three strategies: 1) Separate per-band reconstruction with no shared structure. 2) Splitting optimization, in which we first optimize RGB geometry, copy it, and then fit each new band to the model by optimizing both geometry and band representation. 3) Joint, in which the modalities are jointly optimized, optionally with an initial RGB-only phase. We showcase through quantitative metrics and qualitative novel-view renderings on multi-spectral datasets the effectiveness of our dedicated optimized Joint strategy, increasing overall spectral reconstruction as well as enhancing RGB results through spectral cross-talk. We therefore suggest integrating multi-spectral data directly into the spherical harmonics color components to compactly model each Gaussian's multi-spectral reflectance. Moreover, our analysis reveals several key trade-offs in when and how to introduce spectral bands during optimization, offering practical insights for robust multi-modal 3DGS reconstruction.

cs.CV↗

arXiv2025

Physically Based Neural LiDAR Resimulation

Methods for Novel View Synthesis (NVS) have recently found traction in the field of LiDAR simulation and large-scale 3D scene reconstruction. While solutions for faster rendering or handling dynamic scenes have been proposed, LiDAR specific effects remain insufficiently addressed. By explicitly modeling sensor characteristics such as rolling shutter, laser power variations, and intensity falloff, our method achieves more accurate LiDAR simulation compared to existing techniques. We demonstrate the effectiveness of our approach through quantitative and qualitative comparisons with state-of-the-art methods, as well as ablation studies that highlight the importance of each sensor model component. Beyond that, we show that our approach exhibits advanced resimulation capabilities, such as generating high resolution LiDAR scans in the camera perspective. Our code and the resulting dataset are available at https://github.com/richardmarcus/PBNLiDAR.

cs.RO↗

arXiv2025

Multi-Spectral Gaussian Splatting with Neural Color Representation

We present MS-Splatting -- a multi-spectral 3D Gaussian Splatting (3DGS) framework that is able to generate multi-view consistent novel views from images of multiple, independent cameras with different spectral domains. In contrast to previous approaches, our method does not require cross-modal camera calibration and is versatile enough to model a variety of different spectra, including thermal and near-infra red, without any algorithmic changes. Unlike existing 3DGS-based frameworks that treat each modality separately (by optimizing per-channel spherical harmonics) and therefore fail to exploit the underlying spectral and spatial correlations, our method leverages a novel neural color representation that encodes multi-spectral information into a learned, compact, per-splat feature embedding. A shallow multi-layer perceptron (MLP) then decodes this embedding to obtain spectral color values, enabling joint learning of all bands within a unified representation. Our experiments show that this simple yet effective strategy is able to improve multi-spectral rendering quality, while also leading to improved per-spectra rendering quality over state-of-the-art methods. We demonstrate the effectiveness of this new technique in agricultural applications to render vegetation indices, such as normalized difference vegetation index (NDVI).

cs.GR↗

arXiv2025

FruitNeRF++: A Generalized Multi-Fruit Counting Method Utilizing Contrastive Learning and Neural Radiance Fields

We introduce FruitNeRF++, a novel fruit-counting approach that combines contrastive learning with neural radiance fields to count fruits from unstructured input photographs of orchards. Our work is based on FruitNeRF, which employs a neural semantic field combined with a fruit-specific clustering approach. The requirement for adaptation for each fruit type limits the applicability of the method, and makes it difficult to use in practice. To lift this limitation, we design a shape-agnostic multi-fruit counting framework, that complements the RGB and semantic data with instance masks predicted by a vision foundation model. The masks are used to encode the identity of each fruit as instance embeddings into a neural instance field. By volumetrically sampling the neural fields, we extract a point cloud embedded with the instance features, which can be clustered in a fruit-agnostic manner to obtain the fruit count. We evaluate our approach using a synthetic dataset containing apples, plums, lemons, pears, peaches, and mangoes, as well as a real-world benchmark apple dataset. Our results demonstrate that FruitNeRF++ is easier to control and compares favorably to other state-of-the-art methods.

cs.CV↗

arXiv2025

Synth It Like KITTI: Synthetic Data Generation for Object Detection in Driving Scenarios

An important factor in advancing autonomous driving systems is simulation. Yet, there is rather small progress for transferability between the virtual and real world. We revisit this problem for 3D object detection on LiDAR point clouds and propose a dataset generation pipeline based on the CARLA simulator. Utilizing domain randomization strategies and careful modeling, we are able to train an object detector on the synthetic data and demonstrate strong generalization capabilities to the KITTI dataset. Furthermore, we compare different virtual sensor variants to gather insights, which sensor attributes can be responsible for the prevalent domain gap. Finally, fine-tuning with a small portion of real data almost matches the baseline and with the full training set slightly surpasses it.

cs.CV↗

arXiv2024

MAROON: A Dataset for the Joint Characterization of Near-Field High-Resolution Radio-Frequency and Optical Depth Imaging Techniques

Utilizing the complementary strengths of wavelength-specific range or depth sensors is crucial for robust computer-assisted tasks such as autonomous driving. Despite this, there is still little research done at the intersection of optical depth sensors and radars operating close range, where the target is decimeters away from the sensors. Together with a growing interest in high-resolution imaging radars operating in the near field, the question arises how these sensors behave in comparison to their traditional optical counterparts. In this work, we take on the unique challenge of jointly characterizing depth imagers from both, the optical and radio-frequency domain using a multimodal spatial calibration. We collect data from four depth imagers, with three optical sensors of varying operation principle and an imaging radar. We provide a comprehensive evaluation of their depth measurements with respect to distinct object materials, geometries, and object-to-sensor distances. Specifically, we reveal scattering effects of partially transmissive materials and investigate the response of radio-frequency signals. All object measurements will be made public in form of a multimodal dataset, called MAROON.

eess.IV↗

arXiv2024

VR-Splatting: Foveated Radiance Field Rendering via 3D Gaussian Splatting and Neural Points

Recent advances in novel view synthesis have demonstrated impressive results in fast photorealistic scene rendering through differentiable point rendering, either via Gaussian Splatting (3DGS) [Kerbl and Kopanas et al. 2023] or neural point rendering [Aliev et al. 2020]. Unfortunately, these directions require either a large number of small Gaussians or expensive per-pixel post-processing for reconstructing fine details, which negatively impacts rendering performance. To meet the high performance demands of virtual reality (VR) systems, primitive or pixel counts therefore must be kept low, affecting visual quality. In this paper, we propose a novel hybrid approach based on foveated rendering as a promising solution that combines the strengths of both point rendering directions regarding performance sweet spots. Analyzing the compatibility with the human visual system, we find that using a low-detailed, few primitive smooth Gaussian representation for the periphery is cheap to compute and meets the perceptual demands of peripheral vision. For the fovea only, we use neural points with a convolutional neural network for the small pixel footprint, which provides sharp, detailed output within the rendering budget. This combination also allows for synergistic method accelerations with point occlusion culling and reducing the demands on the neural network. Our evaluation confirms that our approach increases sharpness and details compared to a standard VR-ready 3DGS configuration, and participants of a user study overwhelmingly preferred our method. Our system meets the necessary performance requirements for real-time VR interactions, ultimately enhancing the user's immersive experience. The project page can be found at: https://lfranke.github.io/vr_splatting

cs.CV↗

arXiv2024

Efficient Perspective-Correct 3D Gaussian Splatting Using Hybrid Transparency

3D Gaussian Splats (3DGS) have proven a versatile rendering primitive, both for inverse rendering as well as real-time exploration of scenes. In these applications, coherence across camera frames and multiple views is crucial, be it for robust convergence of a scene reconstruction or for artifact-free fly-throughs. Recent work started mitigating artifacts that break multi-view coherence, including popping artifacts due to inconsistent transparency sorting and perspective-correct outlines of (2D) splats. At the same time, real-time requirements forced such implementations to accept compromises in how transparency of large assemblies of 3D Gaussians is resolved, in turn breaking coherence in other ways. In our work, we aim at achieving maximum coherence, by rendering fully perspective-correct 3D Gaussians while using a high-quality approximation of accurate blending, hybrid transparency, on a per-pixel level, in order to retain real-time frame rates. Our fast and perspectively accurate approach for evaluation of 3D Gaussians does not require matrix inversions, thereby ensuring numerical stability and eliminating the need for special handling of degenerate splats, and the hybrid transparency formulation for blending maintains similar quality as fully resolved per-pixel transparencies at a fraction of the rendering costs. We further show that each of these two components can be independently integrated into Gaussian splatting systems. In combination, they achieve up to 2$\times$ higher frame rates, 2$\times$ faster optimization, and equal or better image quality with fewer rendering artifacts compared to traditional 3DGS on common benchmarks.

cs.GR↗

arXiv2024

Refinement of Monocular Depth Maps via Multi-View Differentiable Rendering

Accurate depth estimation is at the core of many applications in computer graphics, vision, and robotics. Current state-of-the-art monocular depth estimators, trained on extensive datasets, generalize well but lack 3D consistency needed for many applications. In this paper, we combine the strength of those generalizing monocular depth estimation techniques with multi-view data by framing this as an analysis-by-synthesis optimization problem to lift and refine such relative depth maps to accurate error-free depth maps. After an initial global scale estimation through structure-from-motion point clouds, we further refine the depth map through optimization enforcing multi-view consistency via photometric and geometric losses with differentiable rendering of the meshed depth map. In a two-stage optimization, scaling is further refined first, and afterwards artifacts and errors in the depth map are corrected via nearby-view photometric supervision. Our evaluation shows that our method is able to generate detailed, high-quality, view consistent, accurate depth maps, also in challenging indoor scenarios, and outperforms state-of-the-art multi-view depth reconstruction approaches on such datasets. Project page and source code can be found at https://lorafib.github.io/ref_depth/.

cs.CV↗

arXiv2024

End-to-end learned Lossy Dynamic Point Cloud Attribute Compression

Recent advancements in point cloud compression have primarily emphasized geometry compression while comparatively fewer efforts have been dedicated to attribute compression. This study introduces an end-to-end learned dynamic lossy attribute coding approach, utilizing an efficient high-dimensional convolution to capture extensive inter-point dependencies. This enables the efficient projection of attribute features into latent variables. Subsequently, we employ a context model that leverage previous latent space in conjunction with an auto-regressive context model for encoding the latent tensor into a bitstream. Evaluation of our method on widely utilized point cloud datasets from the MPEG and Microsoft demonstrates its superior performance compared to the core attribute compression module Region-Adaptive Hierarchical Transform method from MPEG Geometry Point Cloud Compression with 38.1% Bjontegaard Delta-rate saving in average while ensuring a low-complexity encoding/decoding.

eess.IV↗

arXiv2024

FruitNeRF: A Unified Neural Radiance Field based Fruit Counting Framework

We introduce FruitNeRF, a unified novel fruit counting framework that leverages state-of-the-art view synthesis methods to count any fruit type directly in 3D. Our framework takes an unordered set of posed images captured by a monocular camera and segments fruit in each image. To make our system independent of the fruit type, we employ a foundation model that generates binary segmentation masks for any fruit. Utilizing both modalities, RGB and semantic, we train a semantic neural radiance field. Through uniform volume sampling of the implicit Fruit Field, we obtain fruit-only point clouds. By applying cascaded clustering on the extracted point cloud, our approach achieves precise fruit count.The use of neural radiance fields provides significant advantages over conventional methods such as object tracking or optical flow, as the counting itself is lifted into 3D. Our method prevents double counting fruit and avoids counting irrelevant fruit.We evaluate our methodology using both real-world and synthetic datasets. The real-world dataset consists of three apple trees with manually counted ground truths, a benchmark apple dataset with one row and ground truth fruit location, while the synthetic dataset comprises various fruit types including apple, plum, lemon, pear, peach, and mango.Additionally, we assess the performance of fruit counting using the foundation model compared to a U-Net.

cs.CV↗

arXiv2024

An Efficient yet High-Performance Method for Precise Radar-Based Imaging of Human Hand Poses

Contactless hand pose estimation requires sensors that provide precise spatial information and low computational complexity for real-time processing. Unlike vision-based systems, radar offers lighting independence and direct motion assessments. Yet, there is limited research balancing real-time constraints, suitable frame rates for motion evaluations, and the need for precise 3D data. To address this, we extend the ultra-efficient two-tone hand imaging method from our prior work to a three-tone approach. Maintaining high frame rates and real-time constraints, this approach significantly enhances reconstruction accuracy and precision. We assess these measures by evaluating reconstruction results for different hand poses obtained by an imaging radar. Accuracy is assessed against ground truth from a spatially calibrated photogrammetry setup, while precision is measured using 3D-printed hand poses. The results emphasize the method's great potential for future radar-based hand sensing.

eess.SP↗

arXiv2024

INPC: Implicit Neural Point Clouds for Radiance Field Rendering

We introduce a new approach for reconstruction and novel view synthesis of unbounded real-world scenes. In contrast to previous methods using either volumetric fields, grid-based models, or discrete point cloud proxies, we propose a hybrid scene representation, which implicitly encodes the geometry in a continuous octree-based probability field and view-dependent appearance in a multi-resolution hash grid. This allows for extraction of arbitrary explicit point clouds, which can be rendered using rasterization. In doing so, we combine the benefits of both worlds and retain favorable behavior during optimization: Our novel implicit point cloud representation and differentiable bilinear rasterizer enable fast rendering while preserving the fine geometric detail captured by volumetric neural fields. Furthermore, this representation does not depend on priors like structure-from-motion point clouds. Our method achieves state-of-the-art image quality on common benchmarks. Furthermore, we achieve fast inference at interactive frame rates, and can convert our trained model into a large, explicit point cloud to further enhance performance.

cs.CV↗

arXiv2024

Automatic Spatial Calibration of Near-Field MIMO Radar With Respect to Optical Depth Sensors

Despite an emerging interest in MIMO radar, the utilization of its complementary strengths in combination with optical depth sensors has so far been limited to far-field applications, due to the challenges that arise from mutual sensor calibration in the near field. In fact, most related approaches in the autonomous industry propose target-based calibration methods using corner reflectors that have proven to be unsuitable for the near field. In contrast, we propose a novel, joint calibration approach for optical RGB-D sensors and MIMO radars that is designed to operate in the radar's near-field range, within decimeters from the sensors. Our pipeline consists of a bespoke calibration target, allowing for automatic target detection and localization, followed by the spatial calibration of the two sensor coordinate systems through target registration. We validate our approach using two different depth sensing technologies from the optical domain. The experiments show the efficiency and accuracy of our calibration for various target displacements, as well as its robustness of our localization in terms of signal ambiguities.

cs.RO↗