arXiv subjects

Andreas Kolb

Publications and source records attributed to Andreas Kolb.

At least 19 recordsLinked to original sources

arXiv2026

Smart target point control for Gaussian Splatting methods

Standard Gaussian splatting methods rely on heuristic densification and pruning to adaptively allocate primitives during training, and the resulting Gaussian count strongly influences both reconstruction quality and runtime. This makes comparisons across methods fragile: improvements can stem from higher representational capacity rather than algorithmic design. A common and naive workaround for this is hard-stopping or budgeting densification/pruning once a target count is reached, which biases training because different methods hit the cap at different times, yielding non-uniform densify/prune exposure across views and uneven point distributions. We propose a target point control scheme that preserves the standard densification window and cadence, but adjusts only the existing densification and opacity-culling hyper-parameters to track a quadratic target count trajectory. This quota-governor reaches the desired count by 15k iterations without abrupt cutoffs, ensuring that all methods and views receive equal densification and pruning cycles, enabling fairer, capacity-matched evaluation.

cs.GR↗

arXiv2025

Neural Atlas Graphs for Dynamic Scene Decomposition and Editing

Learning editable high-resolution scene representations for dynamic scenes is an open problem with applications across the domains from autonomous driving to creative editing - the most successful approaches today make a trade-off between editability and supporting scene complexity: neural atlases represent dynamic scenes as two deforming image layers, foreground and background, which are editable in 2D, but break down when multiple objects occlude and interact. In contrast, scene graph models make use of annotated data such as masks and bounding boxes from autonomous-driving datasets to capture complex 3D spatial relationships, but their implicit volumetric node representations are challenging to edit view-consistently. We propose Neural Atlas Graphs (NAGs), a hybrid high-resolution scene representation, where every graph node is a view-dependent neural atlas, facilitating both 2D appearance editing and 3D ordering and positioning of scene elements. Fit at test-time, NAGs achieve state-of-the-art quantitative results on the Waymo Open Dataset - by 5 dB PSNR increase compared to existing methods - and make environmental editing possible in high resolution and visual quality - creating counterfactual driving scenarios with new backgrounds and edited vehicle appearance. We find that the method also generalizes beyond driving scenes and compares favorably - by more than 7 dB in PSNR - to recent matting and video editing baselines on the DAVIS video dataset with a diverse set of human and animal-centric scenes. Project Page: https://princeton-computational-imaging.github.io/nag/

cs.GR↗

arXiv2025

Solving Boundary Handling Analytically in Two Dimensions for Smoothed Particle Hydrodynamics

We present a fully analytic approach for evaluating boundary integrals in two dimensions for Smoothed Particle Hydrodynamics (SPH). Conventional methods often rely on boundary particles or wall re-normalization approaches derived from applying the divergence theorem, whereas our method directly evaluates the area integrals for SPH kernels and gradients over triangular boundaries. This direct integration strategy inherently accommodates higher-order boundary conditions, such as piecewise cubic fields defined via Finite Element stencils, enabling analytic and flexible coupling with mesh-based solvers. At the core of our approach is a general solution for compact polynomials of arbitrary degree over triangles by decomposing the boundary elements into elementary integrals that can be solved with closed-form solutions. We provide a complete, closed-form solution for these generalized integrals, derived by relating the angular components to Chebyshev polynomials and solving the resulting radial integral via a numerically stable evaluation of the Gaussian hypergeometric function $_2F_1$. Our solution is robust and adaptable and works regardless of triangle geometries and kernel functions. We validate the accuracy against high-precision numerical quadrature rules, as well as in problems with known exact solutions. We provide an open-source implementation of our general solution using differentiable programming to facilitate the adoption of our approach to SPH and other contexts that require analytic integration over polygonal domains. Our analytic solution outperforms existing numerical quadrature rules for this problem by up to five orders of magnitude, for integrals and their gradients, while providing a flexible framework to couple arbitrary triangular meshes analytically to Lagrangian schemes, building a strong foundation for addressing several grand challenges in SPH and beyond.

math.NA↗

arXiv2025

Tomato Multi-Angle Multi-Pose Dataset for Fine-Grained Phenotyping

Observer bias and inconsistencies in traditional plant phenotyping methods limit the accuracy and reproducibility of fine-grained plant analysis. To overcome these challenges, we developed TomatoMAP, a comprehensive dataset for Solanum lycopersicum using an Internet of Things (IoT) based imaging system with standardized data acquisition protocols. Our dataset contains 64,464 RGB images that capture 12 different plant poses from four camera elevation angles. Each image includes manually annotated bounding boxes for seven regions of interest (ROIs), including leaves, panicle, batch of flowers, batch of fruits, axillary shoot, shoot and whole plant area, along with 50 fine-grained growth stage classifications based on the BBCH scale. Additionally, we provide 3,616 high-resolution image subset with pixel-wise semantic and instance segmentation annotations for fine-grained phenotyping. We validated our dataset using a cascading model deep learning framework combining MobileNetv3 for classification, YOLOv11 for object detection, and MaskRCNN for segmentation. Through AI vs. Human analysis involving five domain experts, we demonstrate that the models trained on our dataset achieve accuracy and speed comparable to the experts. Cohen's Kappa and inter-rater agreement heatmap confirm the reliability of automated fine-grained phenotyping using our approach.

cs.CV↗

arXiv2023

Decoupled Boundary Handling in SPH

Particle-based boundary representations are frequently used in Smoothed Particle Hydrodynamics (SPH) due to their simple integration into fluid solvers. Commonly, incompressible fluid solvers estimate the current density and corresponding forces in case the current density exceeds the rest density to push fluid particles apart. Close to the boundary, the calculation of the fluid particles' density involves both, neighboring fluid and neighboring boundary particles, yielding an overestimation of density, and, subsequently, wrong pressure forces and wrong velocities leading to the disturbed fluid particles' behavior in the vicinity of the boundary. In this paper, we present a detailed explanation of this disturbed fluid particle behavior, which is mainly due to the combined or coupled handling of the fluid-fluid particle and the fluid-boundary particle interaction. We propose the decoupled handling of both interaction types, leading to two densities for a given fluid particle, i.e., fluid-induced density and boundary-induced density. In our approach, we alternately apply the corresponding fluid-induced and boundary-induced forces during pressure estimation. This separation avoids force overestimation and reduces unintended fluid dynamics near the boundary, as well as a consistent fluid-boundary distance across different fluid amounts and different particle-based boundary handling methods. We compare our method with two regular state-of-the-art methods in different experiments and show how our method handles detailed boundary shapes.

cs.GR↗

arXiv2023

Neural Volume Super-Resolution

Neural volumetric representations have become a widely adopted model for radiance fields in 3D scenes. These representations are fully implicit or hybrid function approximators of the instantaneous volumetric radiance in a scene, which are typically learned from multi-view captures of the scene. We investigate the new task of neural volume super-resolution - rendering high-resolution views corresponding to a scene captured at low resolution. To this end, we propose a neural super-resolution network that operates directly on the volumetric representation of the scene. This approach allows us to exploit an advantage of operating in the volumetric domain, namely the ability to guarantee consistent super-resolution across different viewing directions. To realize our method, we devise a novel 3D representation that hinges on multiple 2D feature planes. This allows us to super-resolve the 3D scene representation by applying 2D convolutional networks on the 2D feature planes. We validate the proposed method by super-resolving multi-view consistent views on a diverse set of unseen 3D scenes, confirming qualitative and quantitatively favorable quality over existing approaches.

cs.CV↗

arXiv2023

Differentiable Sensor Layouts for End-to-End Learning of Task-Specific Camera Parameters

The success of deep learning is frequently described as the ability to train all parameters of a network on a specific application in an end-to-end fashion. Yet, several design choices on the camera level, including the pixel layout of the sensor, are considered as pre-defined and fixed, and high resolution, regular pixel layouts are considered to be the most generic ones in computer vision and graphics, treating all regions of an image as equally important. While several works have considered non-uniform, \eg, hexagonal or foveated, pixel layouts in hardware and image processing, the layout has not been integrated into the end-to-end learning paradigm so far. In this work, we present the first truly end-to-end trained imaging pipeline that optimizes the size and distribution of pixels on the imaging sensor jointly with the parameters of a given neural network on a specific task. We derive an analytic, differentiable approach for the sensor layout parameterization that allows for task-specific, local varying pixel resolutions. We present two pixel layout parameterization functions: rectangular and curvilinear grid shapes that retain a regular topology. We provide a drop-in module that approximates sensor simulation given existing high-resolution images to directly connect our method with existing deep learning models. We show that network predictions benefit from learnable pixel layouts for two different downstream tasks, classification and semantic segmentation.

cs.CV↗

arXiv2022

Glyph from Icon -- Automated Generation of Metaphoric Glyphs

Metaphoric glyphs enhance the readability and learnability of abstract glyphs used for the visualization of quantitative multidimensional data by building upon graphical entities that are intuitively related to the underlying problem domain. Their construction is, however, a predominantly manual process. In this paper, we introduce the Glyph-from-Icon (GfI) approach that allows the automated generation of metaphoric glyphs from user specified icons. Our approach modifies the icon's visual appearance using up to seven quantifiable visual variables, three of which manipulate its geometry while four affect its color. Depending on the visualization goal, specific combinations of these visual variables define the glyphs's variables used for data encoding. Technically, we propose a diffusion-curve based parametric icon representation, which comprises the degrees-of-freedom related to the geometric and color-based visual variables. Moreover, we extend our GfI approach to achieve scalability of the generated glyphs. Based on a user study we evaluate the perception of the glyph's main variables, i.e., amplitude and frequency of geometric and color modulation, as function of the stimuli and deduce functional relations as well as quantization levels to achieve perceptual monotonicity and readability. Finally, we propose a robustly perceivable combination of visual variables, which we apply to the visualization of COVID-19 data.

cs.HC↗

arXiv2021

Light Field Implicit Representation for Flexible Resolution Reconstruction

Inspired by the recent advances in implicitly representing signals with trained neural networks, we aim to learn a continuous representation for narrow-baseline 4D light fields. We propose an implicit representation model for 4D light fields which is conditioned on a sparse set of input views. Our model is trained to output the light field values for a continuous range of query spatio-angular coordinates. Given a sparse set of input views, our scheme can super-resolve the input in both spatial and angular domains by flexible factors. consists of a feature extractor and a decoder which are trained on a dataset of light field patches. The feature extractor captures per-pixel features from the input views. These features can be resized to a desired spatial resolution and fed to the decoder along with the query coordinates. This formulation enables us to reconstruct light field views at any desired spatial and angular resolution. Additionally, our network can handle scenarios in which input views are either of low-resolution or with missing pixels. Experiments show that our method achieves state-of-the-art performance for the task of view synthesis while being computationally fast.

cs.CV↗

arXiv2020

A Generative Model for Generic Light Field Reconstruction

Recently deep generative models have achieved impressive progress in modeling the distribution of training data. In this work, we present for the first time a generative model for 4D light field patches using variational autoencoders to capture the data distribution of light field patches. We develop a generative model conditioned on the central view of the light field and incorporate this as a prior in an energy minimization framework to address diverse light field reconstruction tasks. While pure learning-based approaches do achieve excellent results on each instance of such a problem, their applicability is limited to the specific observation model they have been trained on. On the contrary, our trained light field generative model can be incorporated as a prior into any model-based optimization approach and therefore extend to diverse reconstruction tasks including light field view synthesis, spatial-angular super resolution and reconstruction from coded projections. Our proposed method demonstrates good reconstruction, with performance approaching end-to-end trained networks, while outperforming traditional model-based approaches on both synthetic and real scenes. Furthermore, we show that our approach enables reliable light field recovery despite distortions in the input.

eess.IV↗

arXiv2020

Designing technology, developing theory. Towards a symmetrical approach

We focus on collaborative activities that engage computer graphics designers and social scientists in systems design processes. Our conceptual symmetrical account of technology design and theory development is elaborated as a mode of mutual engagement occurring in an interdisciplinary trading zone, where neither discipline is placed at the service of the other, and nor do disciplinary boundaries dissolve. To this end, we draw on analyses of mutual engagements between computer and social scientists stemming from the fields of computer-supported cooperative work (CSCW), human-computer interaction (HCI), and science and technology studies (STS). We especially build on theoretical work in STS concerning information technology (IT) in health care and extend recent contributions from STS with respect to the modes of engagement and trading zones between computer and social sciences. We conceive participative digital systems design as a form of inquiry for the analysis of cooperative work settings, particularly when social science becomes part of design processes. We illustrate our conceptual approach using data from an interdisciplinary project involving computer graphics designers, sociologists, and neurosurgeons with the aim of developing patient-centered visualizations for clinical cooperation on a hospital ward.

cs.CY↗

arXiv2020

On-Body Visualization of Patient Data for Cooperative Tasks

Electronic health records (EHR) systematically represent patient data in digital form. However, text and visualization based EHR systems are poorly integrated in the hospital workflow due to their complex and rather non-intuitive access structure. This is especially disadvantageous in clinical cooperative situations that require an efficient, task specific information transfer. In this paper we introduce a novel concept of anatomically integrated in-place visualization designed to engage with cooperative tasks on a neurosurgical ward. Based on the findings of our field studies and the derived design goals, we propose an approach that follows a visual tradition in medicine, which is tightly related with anatomy, by using a virtual patient's body as spatial representation of visually encoded abstract medical data. More specifically, we provide a generic set of formal requirements for these kinds of in-place visualizations, we apply these requirements in order to achieve a specific visualization of neurological symptoms related to the differential diagnosis of spinal disc herniation, and we present a prototypical implementation of the visualization concept on a mobile device. Moreover, we discuss various challenges related to visual encoding and visibility of the body model components. Finally, the prototype is evaluated by 10 neurosurgeons, who assess the validity and the further potential of the proposed approach.

cs.HC↗

arXiv2019

Training Auto-encoder-based Optimizers for Terahertz Image Reconstruction

Terahertz (THz) sensing is a promising imaging technology for a wide variety of different applications. Extracting the interpretable and physically meaningful parameters for such applications, however, requires solving an inverse problem in which a model function determined by these parameters needs to be fitted to the measured data. Since the underlying optimization problem is nonconvex and very costly to solve, we propose learning the prediction of suitable parameters from the measured data directly. More precisely, we develop a model-based autoencoder in which the encoder network predicts suitable parameters and the decoder is fixed to a physically meaningful model function, such that we can train the encoding network in an unsupervised way. We illustrate numerically that the resulting network is more than 140 times faster than classical optimization techniques while making predictions with only slightly higher objective values. Using such predictions as starting points of local optimization techniques allows us to converge to better local minima about twice as fast as optimization without the network-based initialization.

cs.CV↗

arXiv2019

Progressive Refinement Imaging

This paper presents a novel technique for progressive online integration of uncalibrated image sequences with substantial geometric and/or photometric discrepancies into a single, geometrically and photometrically consistent image. Our approach can handle large sets of images, acquired from a nearly planar or infinitely distant scene at different resolutions in object domain and under variable local or global illumination conditions. It allows for efficient user guidance as its progressive nature provides a valid and consistent reconstruction at any moment during the online refinement process. Our approach avoids global optimization techniques, as commonly used in the field of image refinement, and progressively incorporates new imagery into a dynamically extendable and memory-efficient Laplacian pyramid. Our image registration process includes a coarse homography and a local refinement stage using optical flow. Photometric consistency is achieved by retaining the photometric intensities given in a reference image, while it is being refined. Globally blurred imagery and local geometric inconsistencies due to, e.g. motion are detected and removed prior to image fusion. We demonstrate the quality and robustness of our approach using several image and video sequences, including handheld acquisition with mobile phones and zooming sequences with consumer cameras.

cs.GR↗

arXiv2019

Computational Image Enhancement for Frequency Modulated Continuous Wave (FMCW) THz Image

In this paper, a novel method to enhance Frequency Modulated Continuous Wave (FMCW) THz imaging resolution beyond its diffraction limit is proposed. Our method comprises two stages. Firstly, we reconstruct the signal in depth-direction using a sinc-envelope, yielding a significant improvement in depth estimation and signal parameter extraction. The resulting high precision depth estimate is used to deduce an accurate reflection intensity THz image. This image is fed in the second stage of our method to a 2D blind deconvolution procedure, adopted to enhance the lateral THz image resolution beyond the diffraction limit. Experimental data acquired with a FMCW system operating at 577 GHz with a bandwidth of 126 GHz shows that the proposed method enhances the lateral resolution by a factor of 2.29 to 346.2um with respect to the diffraction limit. The depth accuracy is 91um. Interestingly, the lateral resolution enhancement achieved with this blind deconvolution concept leads to better results in comparison to conventional gaussian deconvolution. Experimental data on a PCB resolution target is presented, in order to quantify the resolution enhancement and to compare the performance with established image enhancement approaches. The presented technique allows exposure of the interwoven fibre reinforced embedded structures of the PCB test sample.

eess.IV↗

arXiv2019

A Bit Too Much? High Speed Imaging from Sparse Photon Counts

Recent advances in photographic sensing technologies have made it possible to achieve light detection in terms of a single photon. Photon counting sensors are being increasingly used in many diverse applications. We address the problem of jointly recovering spatial and temporal scene radiance from very few photon counts. Our ConvNet-based scheme effectively combines spatial and temporal information present in measurements to reduce noise. We demonstrate that using our method one can acquire videos at a high frame rate and still achieve good quality signal-to-noise ratio. Experiments show that the proposed scheme performs quite well in different challenging scenarios while the existing approaches are unable to handle them.

cs.CV↗

arXiv2018

Supervised Deep Kriging for Single-Image Super-Resolution

We propose a novel single-image super-resolution approach based on the geostatistical method of kriging. Kriging is a zero-bias minimum-variance estimator that performs spatial interpolation based on a weighted average of known observations. Rather than solving for the kriging weights via the traditional method of inverting covariance matrices, we propose a supervised form in which we learn a deep network to generate said weights. We combine the kriging weight generation and kriging process into a joint network that can be learned end-to-end. Our network achieves competitive super-resolution results as other state-of-the-art methods. In addition, since the super-resolution process follows a known statistical framework, we are able to estimate bias and variance, something which is rarely possible for other deep networks.

cs.CV↗

arXiv2018

Residuum-Condition Diagram and Reduction of Over-Complete Endmember-Sets

Extracting reference spectra, or endmembers (EMs) from a given multi- or hyperspectral image, as well as estimating the size of the EM set, plays an important role in multispectral image processing. In this paper, we present condition-residuum-diagrams. By plotting the residuum resulting from the unmixing and reconstruction and the condition number of various EM sets, the resulting diagram provides insight into the behavior of the spectral unmixing under a varying amount of endmembers (EMs). Furthermore, we utilize condition-residuum-diagrams to realize an EM reduction algorithm that starts with an initially extracted, over-complete EM set. An over-complete EM set commonly exhibits a good unmixing result, i.e. a lower reconstruction residuum, but due to its partial redundancy, the unmixing gets numerically unstable, i.e. the unmixed abundances values are less reliable. Our greedy reduction scheme improves the EM set by reducing the condition number, i.e. enhancing the set's stability, while keeping the reconstruction error as low as possible. The resulting set sequence gives hint to the optimal EM set and its size. We demonstrate the benefit of our condition-residuum-diagram and reduction scheme on well-studied datasets with known reference EM set sizes for several well-known EE algorithms.

cs.CV↗