SymGS: Leveraging Reflective Symmetries for 3DGS Compression
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (6) · 4503-4510
DOI: 10.1609/aaai.v40i6.42449 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Keshav Gupta und 5 weiteren Autor:innen.
Abstract (Originalsprache)
3D Gaussian Splatting has emerged as a transformative technique in novel view synthesis, primarily due to its high rendering speed and photorealistic fidelity. However, its memory footprint scales rapidly with scene complexity, often reaching several gigabytes. Existing methods address this issue by introducing compression strategies that exploit primitive level redundancy through similarity detection and quantization. We aim to surpass the compression limits of such methods by incorporating symmetry aware techniques, specifically targeting mirror symmetries to eliminate redundant primitives. We propose a novel compression framework, SymGS, introducing learnable mirrors into the scene, thereby eliminating local and global reflective redundancies for compression. Our framework functions as a plug and play enhancement to state of the art compression methods, (e.g. HAC) to achieve further compression. Compared to HAC, we achieve 1.66x compression across benchmark datasets (upto 3x on large scale scenes). On an average, SymGS enables 108x compression over 3DGS, while preserving rendering quality.
Zitieren
APA
Gupta, K., Sanghvi, A., Palley, S. R., Srivastava, A., Sharma, C., & Sharma, A. (2026). SymGS: Leveraging Reflective Symmetries for 3DGS Compression. Proceedings of the AAAI Conference on Artificial Intelligence, 40(6), 4503-4510. https://doi.org/10.1609/aaai.v40i6.42449
Vancouver
Gupta K, Sanghvi A, Palley SR, Srivastava A, Sharma C, Sharma A. SymGS: Leveraging Reflective Symmetries for 3DGS Compression. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(6):4503-4510. doi:10.1609/aaai.v40i6.42449
BibTeX
@article{Gupta2026,
author = {Gupta, Keshav and Sanghvi, Akshat and Palley, Shreyas Reddy and Srivastava, Astitva and Sharma, Charu and Sharma, Avinash},
title = {SymGS: Leveraging Reflective Symmetries for 3DGS Compression},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {6},
pages = {4503--4510},
doi = {10.1609/aaai.v40i6.42449}
}
EUVIMED