-
Fi-Gaussian: Frequency-Aware Implicit Gaussian Splatting for Single Image Dehazing
Authors:
Yuhan Chen,
Ying Fang,
Guofa Li,
Wenxuan Yu,
Yicui Shi,
Kunyang Huang,
Wenbo Chu,
Keqiang Li
Abstract:
Single image dehazing continues to be hindered by the loss of high-frequency details and the difficulty of accurate physical scattering modeling. To address these issues, we propose Fi-Gaussian, a frequency-aware implicit Gaussian splatting network for single image dehazing. Unlike explicit…
▽ More
Single image dehazing continues to be hindered by the loss of high-frequency details and the difficulty of accurate physical scattering modeling. To address these issues, we propose Fi-Gaussian, a frequency-aware implicit Gaussian splatting network for single image dehazing. Unlike explicit rendering methods that rely on 3D point clouds, our method employs implicit Gaussian splatting to adaptively model the underlying distribution of clear images as a continuous representation in 2D feature space. The core of the network is a frequency-aware implicit Gaussian splatting module, which decouples low-frequency structural information and high-frequency texture information in the frequency domain and then performs adaptive Gaussian aggregation with complex-valued weights to recover fine details. In addition, a physics-driven scattering renormalization mechanism is introduced to estimate the transmission map and atmospheric light under the guidance of implicit Gaussian priors. Extensive experiments on multiple benchmark datasets demonstrate that Fi-Gaussian achieves state-of-the-art quantitative performance and produces visually superior dehazed results, validating the effectiveness of implicit Gaussian splatting for low-level vision tasks.
△ Less
Submitted 14 June, 2026;
originally announced June 2026.
-
Gabor Fields: Orientation-Selective Level-of-Detail for Volume Rendering
Authors:
Jorge Condor,
Nicolai Hermann,
Mehmet Ata Yurtsever,
Piotr Didyk
Abstract:
Gaussian-based representations have enabled efficient physically-based volume rendering at a fraction of the memory cost of regular, discrete, voxel-based distributions. However, several remaining issues hamper their widespread use. One of the advantages of classic voxel grids is the ease of constructing hierarchical representations by either storing volumet…
▽ More
Gaussian-based representations have enabled efficient physically-based volume rendering at a fraction of the memory cost of regular, discrete, voxel-based distributions. However, several remaining issues hamper their widespread use. One of the advantages of classic voxel grids is the ease of constructing hierarchical representations by either storing volumetric mipmaps or selectively pruning branches of an already hierarchical voxel grid. Such strategies reduce rendering time and eliminate aliasing when lower levels of detail are required. Constructing similar strategies for Gaussian-based volumes is not trivial. Straightforward solutions, such as prefiltering or computing mipmap-style representations, lead to increased memory requirements or expensive re-fitting of each level separately. Additionally, such solutions do not guarantee a smooth transition between different hierarchy levels. To address these limitations, we propose Gabor Fields, an orientation-selective mixture of Gabor kernels that enables continuous frequency filtering at no cost. The frequency content of the asset is reduced by selectively pruning primitives, directly benefiting rendering performance. Beyond filtering, we demonstrate that stochastically sampling from different frequencies and orientations at each ray recursion enables masking substantial portions of the volume, accelerating ray traversal time in single- and multiple-scattering settings. Furthermore, inspired by procedural volumes, we present an application for efficient design and rendering of procedural clouds as Gabor-noise-modulated Gaussians.
△ Less
Submitted 16 July, 2026; v1 submitted 4 February, 2026;
originally announced February 2026.
-
Light Transport-aware Diffusion Posterior Sampling for Single-View Reconstruction of 3D Volumes
Authors:
Ludwic Leonard,
Nils Thuerey,
Ruediger Westermann
Abstract:
We introduce a single-view reconstruction technique of volumetric fields in which multiple light…
▽ More
We introduce a single-view reconstruction technique of volumetric fields in which multiple light scattering effects are omnipresent, such as in clouds. We model the unknown distribution of volumetric fields using an unconditional diffusion model trained on a novel benchmark dataset comprising 1,000 synthetically simulated volumetric density fields. The neural diffusion model is trained on the latent codes of a novel, diffusion-friendly, monoplanar representation. The generative model is used to incorporate a tailored parametric diffusion posterior sampling technique into different reconstruction tasks. A physically-based differentiable volume renderer is employed to provide gradients with respect to light transport in the latent space. This stands in contrast to classic NeRF approaches and makes the reconstructions better aligned with observed data. Through various experiments, we demonstrate single-view reconstruction of volumetric clouds at a previously unattainable quality.
△ Less
Submitted 28 March, 2025; v1 submitted 9 January, 2025;
originally announced January 2025.
-
Rendering Participating Media Using Path Graphs
Authors:
Becky Hu,
Xi Deng,
Fujun Luan,
Miloš Hašan,
Steve Marschner
Abstract:
Rendering volumetric…
▽ More
Rendering volumetric scattering media, including clouds, fog, smoke, and other complex materials, is crucial for realism in computer graphics. Traditional path tracing, while unbiased, requires many long path samples to converge in scenes with scattering media, and a lot of work is wasted by paths that make a negligible contribution to the image. Methods to make better use of the information learned during path tracing range from photon mapping to radiance caching, but struggle to support the full range of heterogeneous scattering media. This paper introduces a new volumetric rendering algorithm that extends and adapts the previous \emph{path graph} surface rendering algorithm. Our method leverages the information collected through multiple-scattering transport paths to compute lower-noise estimates, increasing computational efficiency by reducing the required sample count. Our key contributions include an extended path graph for participating media and new aggregation and propagation operators for efficient path reuse in volumes. Compared to previous methods, our approach significantly boosts convergence in scenes with challenging volumetric light transport, including heterogeneous media with high scattering albedos and dense, forward-scattering translucent materials, under complex lighting conditions.
△ Less
Submitted 18 April, 2024;
originally announced April 2024.
-
TRIDENT: A Rapid 3D Radiative Transfer Model for Exoplanet Transmission Spectra
Authors:
Ryan J. MacDonald,
Nikole K. Lewis
Abstract:
Transmission spectroscopy is one of the premier methods used to probe the temperature, composition, and cloud properties of exoplanet atmospheres. Recent studies have demonstrated that the multidimensional nature of exoplanet atmospheres -- due to non-uniformities across the day-night transition and between the morning and evening terminators -- can strongly…
▽ More
Transmission spectroscopy is one of the premier methods used to probe the temperature, composition, and cloud properties of exoplanet atmospheres. Recent studies have demonstrated that the multidimensional nature of exoplanet atmospheres -- due to non-uniformities across the day-night transition and between the morning and evening terminators -- can strongly influence transmission spectra. However, the computational demands of 3D radiative transfer techniques have precluded their usage within atmospheric retrievals. Here we introduce TRIDENT, a new 3D radiative transfer model which rapidly computes transmission spectra of exoplanet atmospheres with day-night, morning-evening, and vertical variations in temperature, chemical abundances, and cloud properties. We also derive a general equation for transmission spectra, accounting for 3D atmospheres, refraction, multiple scattering, ingress/egress, grazing transits, stellar heterogeneities, and nightside thermal emission. After introducing TRIDENT's linear algebra-based approach to 3D radiative transfer, we propose new parametric prescriptions for 3D temperature and abundance profiles and 3D clouds. We show that multidimensional transmission spectra exhibit two significant observational signatures: (i) day-night composition gradients alter the relative amplitudes of absorption features; and (ii) morning-evening composition gradients distort the peak-to-wing contrast of absorption features. Finally, we demonstrate that these signatures of multidimensional atmospheres incur residuals > 100 ppm compared to 1D models, rendering them potentially detectable with JWST. TRIDENT's rapid radiative transfer, coupled with parametric multidimensional atmospheres, unlocks the final barrier to 3D atmospheric retrievals.
△ Less
Submitted 31 December, 2021; v1 submitted 10 November, 2021;
originally announced November 2021.
