Learning Representations from 3D Gaussian Splats
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914612889255936 |
|---|---|
| author | Farganus, Julia Żurawicki, Krzysztof Gaweł, Arkadiusz Jakubowska, Weronika Kwaśnicka, Halina |
| author_facet | Farganus, Julia Żurawicki, Krzysztof Gaweł, Arkadiusz Jakubowska, Weronika Kwaśnicka, Halina |
| contents | 3D Gaussian Splatting (3DGS) is a recent approach for scene rendering. Although primarily designed for view synthesis, its potential for scene understanding tasks remains underexplored. In this work, we conduct a comparative evaluation of various geometric deep learning architectures for the classification of 3D scenes represented using Gaussian Splatting. We benchmark point-based and graph-based models across both traditional point cloud datasets and dedicated Gaussian Splatting datasets. Scenes are embedded into latent representations, which are evaluated through end-to-end classification, linear probing, and clustering analysis. Our study provides insight into the suitability of different geometry-aware architectures and input feature configurations for learning effective 3D Gaussian Splat representations. The results highlight consistent differences between architectural families and reveal the impact of Gaussian-specific attributes on the quality of representation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_29549 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Learning Representations from 3D Gaussian Splats Farganus, Julia Żurawicki, Krzysztof Gaweł, Arkadiusz Jakubowska, Weronika Kwaśnicka, Halina Computer Vision and Pattern Recognition 3D Gaussian Splatting (3DGS) is a recent approach for scene rendering. Although primarily designed for view synthesis, its potential for scene understanding tasks remains underexplored. In this work, we conduct a comparative evaluation of various geometric deep learning architectures for the classification of 3D scenes represented using Gaussian Splatting. We benchmark point-based and graph-based models across both traditional point cloud datasets and dedicated Gaussian Splatting datasets. Scenes are embedded into latent representations, which are evaluated through end-to-end classification, linear probing, and clustering analysis. Our study provides insight into the suitability of different geometry-aware architectures and input feature configurations for learning effective 3D Gaussian Splat representations. The results highlight consistent differences between architectural families and reveal the impact of Gaussian-specific attributes on the quality of representation. |
| title | Learning Representations from 3D Gaussian Splats |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2605.29549 |