Neural Surface Priors for Editable Gaussian Splatting
Authors:
Jakub Szymkowiak,
Weronika Jakubowska,
Dawid Malarz,
Weronika Smolak-Dyżewska,
Maciej Zięba,
Przemyslaw Musialski,
Wojtek Pałubicki,
Przemysław Spurek
Abstract:
In computer graphics and vision, recovering easily modifiable scene appearance from image data is crucial for applications such as content creation. We introduce a novel method that integrates 3D Gaussian Splatting with an implicit surface representation, enabling intuitive editing of recovered scenes through mesh manipulation. Starting with a set of input images and camera poses, our approach rec…
▽ More
In computer graphics and vision, recovering easily modifiable scene appearance from image data is crucial for applications such as content creation. We introduce a novel method that integrates 3D Gaussian Splatting with an implicit surface representation, enabling intuitive editing of recovered scenes through mesh manipulation. Starting with a set of input images and camera poses, our approach reconstructs the scene surface using a neural signed distance field. This neural surface acts as a geometric prior guiding the training of Gaussian Splatting components, ensuring their alignment with the scene geometry. To facilitate editing, we encode the visual and geometric information into a lightweight triangle soup proxy. Edits applied to the mesh extracted from the neural surface propagate seamlessly through this intermediate structure to update the recovered appearance. Unlike previous methods relying on the triangle soup proxy representation, our approach supports a wider range of modifications and fully leverages the mesh topology, enabling a more flexible and intuitive editing process. The complete source code for this project can be accessed at: https://github.com/WJakubowska/NeuralSurfacePriors.
△ Less
Submitted 7 February, 2025; v1 submitted 27 November, 2024;
originally announced November 2024.
HOSC: A Periodic Activation Function for Preserving Sharp Features in Implicit Neural Representations
Authors:
Danzel Serrano,
Jakub Szymkowiak,
Przemyslaw Musialski
Abstract:
Recently proposed methods for implicitly representing signals such as images, scenes, or geometries using coordinate-based neural network architectures often do not leverage the choice of activation functions, or do so only to a limited extent. In this paper, we introduce the Hyperbolic Oscillation function (HOSC), a novel activation function with a controllable sharpness parameter. Unlike any pre…
▽ More
Recently proposed methods for implicitly representing signals such as images, scenes, or geometries using coordinate-based neural network architectures often do not leverage the choice of activation functions, or do so only to a limited extent. In this paper, we introduce the Hyperbolic Oscillation function (HOSC), a novel activation function with a controllable sharpness parameter. Unlike any previous activations, HOSC has been specifically designed to better capture sudden changes in the input signal, and hence sharp or acute features of the underlying data, as well as smooth low-frequency transitions. Due to its simplicity and modularity, HOSC offers a plug-and-play functionality that can be easily incorporated into any existing method employing a neural network as a way of implicitly representing a signal. We benchmark HOSC against other popular activations in an array of general tasks, empirically showing an improvement in the quality of obtained representations, provide the mathematical motivation behind the efficacy of HOSC, and discuss its limitations.
△ Less
Submitted 20 January, 2024;
originally announced January 2024.
