Binary-Gaussian: Compact and Progressive Representation for 3D Gaussian Segmentation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909936843227136 |
|---|---|
| author | Yang, An Liu, Chenyu Du, Jun Gao, Jianqing Pan, Jia Hu, Jinshui Yin, Baocai Yin, Bing Liu, Cong |
| author_facet | Yang, An Liu, Chenyu Du, Jun Gao, Jianqing Pan, Jia Hu, Jinshui Yin, Baocai Yin, Bing Liu, Cong |
| contents | 3D Gaussian Splatting (3D-GS) has emerged as an efficient 3D representation and a promising foundation for semantic tasks like segmentation. However, existing 3D-GS-based segmentation methods typically rely on high-dimensional category features, which introduce substantial memory overhead. Moreover, fine-grained segmentation remains challenging due to label space congestion and the lack of stable multi-granularity control mechanisms. To address these limitations, we propose a coarse-to-fine binary encoding scheme for per-Gaussian category representation, which compresses each feature into a single integer via the binary-to-decimal mapping, drastically reducing memory usage. We further design a progressive training strategy that decomposes panoptic segmentation into a series of independent sub-tasks, reducing inter-class conflicts and thereby enhancing fine-grained segmentation capability. Additionally, we fine-tune opacity during segmentation training to address the incompatibility between photometric rendering and semantic segmentation, which often leads to foreground-background confusion. Extensive experiments on multiple benchmarks demonstrate that our method achieves state-of-the-art segmentation performance while significantly reducing memory consumption and accelerating inference. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_00944 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Binary-Gaussian: Compact and Progressive Representation for 3D Gaussian Segmentation Yang, An Liu, Chenyu Du, Jun Gao, Jianqing Pan, Jia Hu, Jinshui Yin, Baocai Yin, Bing Liu, Cong Computer Vision and Pattern Recognition 3D Gaussian Splatting (3D-GS) has emerged as an efficient 3D representation and a promising foundation for semantic tasks like segmentation. However, existing 3D-GS-based segmentation methods typically rely on high-dimensional category features, which introduce substantial memory overhead. Moreover, fine-grained segmentation remains challenging due to label space congestion and the lack of stable multi-granularity control mechanisms. To address these limitations, we propose a coarse-to-fine binary encoding scheme for per-Gaussian category representation, which compresses each feature into a single integer via the binary-to-decimal mapping, drastically reducing memory usage. We further design a progressive training strategy that decomposes panoptic segmentation into a series of independent sub-tasks, reducing inter-class conflicts and thereby enhancing fine-grained segmentation capability. Additionally, we fine-tune opacity during segmentation training to address the incompatibility between photometric rendering and semantic segmentation, which often leads to foreground-background confusion. Extensive experiments on multiple benchmarks demonstrate that our method achieves state-of-the-art segmentation performance while significantly reducing memory consumption and accelerating inference. |
| title | Binary-Gaussian: Compact and Progressive Representation for 3D Gaussian Segmentation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.00944 |