MiraGe: Editable 2D Images using Gaussian Splatting

Joanna Waczynska, Tomasz Szczepanik, Piotr Borycki, Slawomir Tadeja, Thomas Bohné, Przemysław Spurek
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:61868-61884, 2025.

Abstract

Implicit Neural Representations (INRs) approximate discrete data through continuous functions and are commonly used for encoding 2D images. Traditional image-based INRs employ neural networks to map pixel coordinates to RGB values, capturing shapes, colors, and textures within the network’s weights. Recently, GaussianImage has been proposed as an alternative, using Gaussian functions instead of neural networks to achieve comparable quality and compression. Such a solution obtains a quality and compression ratio similar to classical INR models but does not allow image modification. In contrast, our work introduces a novel method, MiraGe, which uses mirror reflections to perceive 2D images in 3D space and employs flat-controlled Gaussians for precise 2D image editing. Our approach improves the rendering quality and allows realistic image modifications, including human-inspired perception of photos in the 3D world. Thanks to modeling images in 3D space, we obtain the illusion of 3D-based modification in 2D images. We also show that our Gaussian representation can be easily combined with a physics engine to produce physics-based modification of 2D images. Consequently, MiraGe allows for better quality than the standard approach and natural modification of 2D images.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-waczynska25a, title = {{M}ira{G}e: Editable 2{D} Images using {G}aussian Splatting}, author = {Waczynska, Joanna and Szczepanik, Tomasz and Borycki, Piotr and Tadeja, Slawomir and Bohn\'{e}, Thomas and Spurek, Przemys{\l}aw}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {61868--61884}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/waczynska25a/waczynska25a.pdf}, url = {https://proceedings.mlr.press/v267/waczynska25a.html}, abstract = {Implicit Neural Representations (INRs) approximate discrete data through continuous functions and are commonly used for encoding 2D images. Traditional image-based INRs employ neural networks to map pixel coordinates to RGB values, capturing shapes, colors, and textures within the network’s weights. Recently, GaussianImage has been proposed as an alternative, using Gaussian functions instead of neural networks to achieve comparable quality and compression. Such a solution obtains a quality and compression ratio similar to classical INR models but does not allow image modification. In contrast, our work introduces a novel method, MiraGe, which uses mirror reflections to perceive 2D images in 3D space and employs flat-controlled Gaussians for precise 2D image editing. Our approach improves the rendering quality and allows realistic image modifications, including human-inspired perception of photos in the 3D world. Thanks to modeling images in 3D space, we obtain the illusion of 3D-based modification in 2D images. We also show that our Gaussian representation can be easily combined with a physics engine to produce physics-based modification of 2D images. Consequently, MiraGe allows for better quality than the standard approach and natural modification of 2D images.} }
Endnote
%0 Conference Paper %T MiraGe: Editable 2D Images using Gaussian Splatting %A Joanna Waczynska %A Tomasz Szczepanik %A Piotr Borycki %A Slawomir Tadeja %A Thomas Bohné %A Przemysław Spurek %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-waczynska25a %I PMLR %P 61868--61884 %U https://proceedings.mlr.press/v267/waczynska25a.html %V 267 %X Implicit Neural Representations (INRs) approximate discrete data through continuous functions and are commonly used for encoding 2D images. Traditional image-based INRs employ neural networks to map pixel coordinates to RGB values, capturing shapes, colors, and textures within the network’s weights. Recently, GaussianImage has been proposed as an alternative, using Gaussian functions instead of neural networks to achieve comparable quality and compression. Such a solution obtains a quality and compression ratio similar to classical INR models but does not allow image modification. In contrast, our work introduces a novel method, MiraGe, which uses mirror reflections to perceive 2D images in 3D space and employs flat-controlled Gaussians for precise 2D image editing. Our approach improves the rendering quality and allows realistic image modifications, including human-inspired perception of photos in the 3D world. Thanks to modeling images in 3D space, we obtain the illusion of 3D-based modification in 2D images. We also show that our Gaussian representation can be easily combined with a physics engine to produce physics-based modification of 2D images. Consequently, MiraGe allows for better quality than the standard approach and natural modification of 2D images.
APA
Waczynska, J., Szczepanik, T., Borycki, P., Tadeja, S., Bohné, T. & Spurek, P.. (2025). MiraGe: Editable 2D Images using Gaussian Splatting. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:61868-61884 Available from https://proceedings.mlr.press/v267/waczynska25a.html.

Related Material