arXiv subjects

Sebastian Weiss

Publications and source records attributed to Sebastian Weiss.

15 recordsLinked to original sources

arXiv2026

FastGHA: Generalized Few-Shot 3D Gaussian Head Avatars with Real-Time Animation

Despite recent progress in 3D Gaussian-based head avatar modeling, efficiently generating high fidelity avatars remains a challenge. Current methods typically rely on extensive multi-view capture setups or monocular videos with per-identity optimization during inference, limiting their scalability and ease of use on unseen subjects. To overcome these efficiency drawbacks, we propose FastGHA, a feed-forward method to generate high-quality Gaussian head avatars from only a few input images while supporting real-time animation. Our approach directly learns a per-pixel Gaussian representation from the input images, and aggregates multi-view information using a transformer-based encoder that fuses image features from both DINOv3 and Stable Diffusion VAE. For real-time animation, we extend the explicit Gaussian representations with per-Gaussian features and introduce a lightweight MLP-based dynamic network to predict 3D Gaussian deformations from expression codes. Furthermore, to enhance geometric smoothness of the 3D head, we employ point maps from a pre-trained large reconstruction model as geometry supervision. Experiments show that our approach significantly outperforms existing methods in both rendering quality and inference efficiency, while supporting real-time dynamic avatar animation.

cs.CV↗

arXiv2026

RelightAnyone: A Generalized Relightable 3D Gaussian Head Model

3D Gaussian Splatting (3DGS) has become a standard approach to reconstruct and render photorealistic 3D head avatars. A major challenge is to relight the avatars to match any scene illumination. For high quality relighting, existing methods require subjects to be captured under complex time-multiplexed illumination, such as one-light-at-a-time (OLAT). We propose a new generalized relightable 3D Gaussian head model that can relight any subject observed in a single- or multi-view images without requiring OLAT data for that subject. Our core idea is to learn a mapping from flat-lit 3DGS avatars to corresponding relightable Gaussian parameters for that avatar. Our model consists of two stages: a first stage that models flat-lit 3DGS avatars without OLAT lighting, and a second stage that learns the mapping to physically-based reflectance parameters for high-quality relighting. This two-stage design allows us to train the first stage across diverse existing multi-view datasets without OLAT lighting ensuring cross-subject generalization, where we learn a dataset-specific lighting code for self-supervised lighting alignment. Subsequently, the second stage can be trained on a significantly smaller dataset of subjects captured under OLAT illumination. Together, this allows our method to generalize well and relight any subject from the first stage as if we had captured them under OLAT lighting. Furthermore, we can fit our model to unseen subjects from as little as a single image, allowing several applications in novel view synthesis and relighting for digital avatars.

cs.CV↗

arXiv2025

ScaffoldAvatar: High-Fidelity Gaussian Avatars with Patch Expressions

Generating high-fidelity real-time animated sequences of photorealistic 3D head avatars is important for many graphics applications, including immersive telepresence and movies. This is a challenging problem particularly when rendering digital avatar close-ups for showing character's facial microfeatures and expressions. To capture the expressive, detailed nature of human heads, including skin furrowing and finer-scale facial movements, we propose to couple locally-defined facial expressions with 3D Gaussian splatting to enable creating ultra-high fidelity, expressive and photorealistic 3D head avatars. In contrast to previous works that operate on a global expression space, we condition our avatar's dynamics on patch-based local expression features and synthesize 3D Gaussians at a patch level. In particular, we leverage a patch-based geometric 3D face model to extract patch expressions and learn how to translate these into local dynamic skin appearance and motion by coupling the patches with anchor points of Scaffold-GS, a recent hierarchical scene representation. These anchors are then used to synthesize 3D Gaussians on-the-fly, conditioned by patch-expressions and viewing direction. We employ color-based densification and progressive training to obtain high-quality results and faster convergence for high resolution 3K training images. By leveraging patch-level expressions, ScaffoldAvatar consistently achieves state-of-the-art performance with visually natural motion, while encompassing diverse facial expressions and styles in real time.

cs.GR↗

arXiv2024

Gaussian Billboards: Expressive 2D Gaussian Splatting with Textures

Gaussian Splatting has recently emerged as the go-to representation for reconstructing and rendering 3D scenes. The transition from 3D to 2D Gaussian primitives has further improved multi-view consistency and surface reconstruction accuracy. In this work we highlight the similarity between 2D Gaussian Splatting (2DGS) and billboards from traditional computer graphics. Both use flat semi-transparent 2D geometry that is positioned, oriented and scaled in 3D space. However 2DGS uses a solid color per splat and an opacity modulated by a Gaussian distribution, where billboards are more expressive, modulating the color with a uv-parameterized texture. We propose to unify these concepts by presenting Gaussian Billboards, a modification of 2DGS to add spatially-varying color achieved using per-splat texture interpolation. The result is a mixture of the two representations, which benefits from both the robust scene optimization power of 2DGS and the expressiveness of texture mapping. We show that our method can improve the sharpness and quality of the scene representation in a wide range of qualitative and quantitative evaluations compared to the original 2DGS implementation.

cs.CV↗

arXiv2024

Monocular Facial Appearance Capture in the Wild

We present a new method for reconstructing the appearance properties of human faces from a lightweight capture procedure in an unconstrained environment. Our method recovers the surface geometry, diffuse albedo, specular intensity and specular roughness from a monocular video containing a simple head rotation in-the-wild. Notably, we make no simplifying assumptions on the environment lighting, and we explicitly take visibility and occlusions into account. As a result, our method can produce facial appearance maps that approach the fidelity of studio-based multi-view captures, but with a far easier and cheaper procedure.

cs.CV↗

arXiv2024

Multimodal Conditional 3D Face Geometry Generation

We present a new method for multimodal conditional 3D face geometry generation that allows user-friendly control over the output identity and expression via a number of different conditioning signals. Within a single model, we demonstrate 3D faces generated from artistic sketches, portrait photos, Canny edges, FLAME face model parameters, 2D face landmarks, or text prompts. Our approach is based on a diffusion process that generates 3D geometry in a 2D parameterized UV domain. Geometry generation passes each conditioning signal through a set of cross-attention layers (IP-Adapter), one set for each user-defined conditioning signal. The result is an easy-to-use 3D face generation tool that produces topology-consistent, high-quality geometry with fine-grain user control.

cs.CV↗

arXiv2023

Artist-Friendly Relightable and Animatable Neural Heads

An increasingly common approach for creating photo-realistic digital avatars is through the use of volumetric neural fields. The original neural radiance field (NeRF) allowed for impressive novel view synthesis of static heads when trained on a set of multi-view images, and follow up methods showed that these neural representations can be extended to dynamic avatars. Recently, new variants also surpassed the usual drawback of baked-in illumination in neural representations, showing that static neural avatars can be relit in any environment. In this work we simultaneously tackle both the motion and illumination problem, proposing a new method for relightable and animatable neural heads. Our method builds on a proven dynamic avatar approach based on a mixture of volumetric primitives, combined with a recently-proposed lightweight hardware setup for relightable neural fields, and includes a novel architecture that allows relighting dynamic neural avatars performing unseen expressions in any environment, even with nearfield illumination and viewpoints.

cs.CV↗

arXiv2021

Fast Neural Representations for Direct Volume Rendering

Despite the potential of neural scene representations to effectively compress 3D scalar fields at high reconstruction quality, the computational complexity of the training and data reconstruction step using scene representation networks limits their use in practical applications. In this paper, we analyze whether scene representation networks can be modified to reduce these limitations and whether such architectures can also be used for temporal reconstruction tasks. We propose a novel design of scene representation networks using GPU tensor cores to integrate the reconstruction seamlessly into on-chip raytracing kernels, and compare the quality and performance of this network to alternative network- and non-network-based compression schemes. The results indicate competitive quality of our design at high compression rates, and significantly faster decoding times and lower memory consumption during data reconstruction. We investigate how density gradients can be computed using the network and show an extension where density, gradient and curvature are predicted jointly. As an alternative to spatial super-resolution approaches for time-varying fields, we propose a solution that builds upon latent-space interpolation to enable random access reconstruction at arbitrary granularity. We summarize our findings in the form of an assessment of the strengths and limitations of scene representation networks \changed{for compression domain volume rendering, and outline future research directions.

cs.GR↗

arXiv2021

Differentiable Direct Volume Rendering

We present a differentiable volume rendering solution that provides differentiability of all continuous parameters of the volume rendering process. This differentiable renderer is used to steer the parameters towards a setting with an optimal solution of a problem-specific objective function. We have tailored the approach to volume rendering by enforcing a constant memory footprint via analytic inversion of the blending functions. This makes it independent of the number of sampling steps through the volume and facilitates the consideration of small-scale changes. The approach forms the basis for automatic optimizations regarding external parameters of the rendering process and the volumetric density field itself. We demonstrate its use for automatic viewpoint selection using differentiable entropy as objective, and for optimizing a transfer function from rendered images of a given volume. Optimization of per-voxel densities is addressed in two different ways: First, we mimic inverse tomography and optimize a 3D density field from images using an absorption model. This simplification enables comparisons with algebraic reconstruction techniques and state-of-the-art differentiable path tracers. Second, we introduce a novel approach for tomographic reconstruction from images using an emission-absorption model with post-shading via an arbitrary transfer function.

cs.GR↗

arXiv2020

Learning Adaptive Sampling and Reconstruction for Volume Visualization

A central challenge in data visualization is to understand which data samples are required to generate an image of a data set in which the relevant information is encoded. In this work, we make a first step towards answering the question of whether an artificial neural network can predict where to sample the data with higher or lower density, by learning of correspondences between the data, the sampling patterns and the generated images. We introduce a novel neural rendering pipeline, which is trained end-to-end to generate a sparse adaptive sampling structure from a given low-resolution input image, and reconstructs a high-resolution image from the sparse set of samples. For the first time, to the best of our knowledge, we demonstrate that the selection of structures that are relevant for the final visual representation can be jointly learned together with the reconstruction of this representation from these structures. Therefore, we introduce differentiable sampling and reconstruction stages, which can leverage back-propagation based on supervised losses solely on the final image. We shed light on the adaptive sampling patterns generated by the network pipeline and analyze its use for volume visualization including isosurface and direct volume rendering.

cs.GR↗

arXiv2019

Sparse Surface Constraints for Combining Physics-based Elasticity Simulation and Correspondence-Free Object Reconstruction

We address the problem to infer physical material parameters and boundary conditions from the observed motion of a homogeneous deformable object via the solution of an inverse problem. Parameters are estimated from potentially unreliable real-world data sources such as sparse observations without correspondences. We introduce a novel Lagrangian-Eulerian optimization formulation, including a cost function that penalizes differences to observations during an optimization run. This formulation matches correspondence-free, sparse observations from a single-view depth sequence with a finite element simulation of deformable bodies. In conjunction with an efficient hexahedral discretization and a stable, implicit formulation of collisions, our method can be used in demanding situation to recover a variety of material parameters, ranging from Young's modulus and Poisson ratio to gravity and stiffness damping, and even external boundaries. In a number of tests using synthetic datasets and real-world measurements, we analyse the robustness of our approach and the convergence behavior of the numerical optimization scheme.

cs.GR↗

arXiv2019

Volumetric Isosurface Rendering with Deep Learning-Based Super-Resolution

Rendering an accurate image of an isosurface in a volumetric field typically requires large numbers of data samples. Reducing the number of required samples lies at the core of research in volume rendering. With the advent of deep learning networks, a number of architectures have been proposed recently to infer missing samples in multi-dimensional fields, for applications such as image super-resolution and scan completion. In this paper, we investigate the use of such architectures for learning the upscaling of a low-resolution sampling of an isosurface to a higher resolution, with high fidelity reconstruction of spatial detail and shading. We introduce a fully convolutional neural network, to learn a latent representation generating a smooth, edge-aware normal field and ambient occlusions from a low-resolution normal and depth field. By adding a frame-to-frame motion loss into the learning stage, the upscaling can consider temporal variations and achieves improved frame-to-frame coherence. We demonstrate the quality of the network for isosurfaces which were never seen during training, and discuss remote and in-situ visualization as well as focus+context visualization as potential applications

cs.GR↗

arXiv2008

Penrose Limits vs String Expansions

We analyse the relation between two a priori quite different expansions of the string equations of motion and constraints in a general curved background, namely one based on the covariant Penrose-Fermi expansion of the metric G_{μν} around a Penrose limit plane wave associated to a null geodesic γ, and the other on the Riemann coordinate expansion in the exact metric G_{μν} of the string embedding variables around the null geodesic γ. Starting with the observation that there is a formal analogy between the exact string equations in a plane wave and the first order string equations in a general background, we show that this analogy becomes exact provided that one chooses the background string configuration to be the null geodesic γitself. We then explore the higher-order correspondence between these two expansions and find that for a general curved background they agree to all orders provided that one works in Fermi coordinates and in the lightcone gauge. Requiring moreover the conformal gauge restricts one to the usual class of (Brinkmann) backgrounds admitting simultaneously the lightcone and the conformal gauge, without further restrictions.

hep-th↗

arXiv2006

Fermi Coordinates and Penrose Limits

We propose a formulation of the Penrose plane wave limit in terms of null Fermi coordinates. This provides a physically intuitive (Fermi coordinates are direct measures of geodesic distance in space-time) and manifestly covariant description of the expansion around the plane wave metric in terms of components of the curvature tensor of the original metric, and generalises the covariant description of the lowest order Penrose limit metric itself, obtained in hep-th/0312029. We describe in some detail the construction of null Fermi coordinates and the corresponding expansion of the metric, and then study various aspects of the higher order corrections to the Penrose limit. In particular, we observe that in general the first-order corrected metric is such that it admits a light-cone gauge description in string theory. We also establish a formal analogue of the Weyl tensor peeling theorem for the Penrose limit expansion in any dimension, and we give a simple derivation of the leading (quadratic) corrections to the Penrose limit of AdS_5 x S^5.

hep-th↗

arXiv2006

Scalar Field Probes of Power-Law Space-Time Singularities

We analyse the effective potential of the scalar wave equation near generic space-time singularities of power-law type (Szekeres-Iyer metrics) and show that the effective potential exhibits a universal and scale invariant leading x^{-2} inverse square behaviour in the ``tortoise coordinate'' x provided that the metrics satisfy the strict Dominant Energy Condition (DEC). This result parallels that obtained in hep-th/0403252 for probes consisting of families of massless particles (null geodesic deviation, a.k.a. the Penrose Limit). The detailed properties of the scalar wave operator depend sensitively on the numerical coefficient of the x^{-2}-term, and as one application we show that timelike singularities satisfying the DEC are quantum mechanically singular in the sense of the Horowitz-Marolf (essential self-adjointness) criterion. We also comment on some related issues like the near-singularity behaviour of the scalar fields permitted by the Friedrichs extension.

hep-th↗