Binary-Gaussian: Compact and Progressive Representation for 3D Gaussian Segmentation

TL;DR

Binary-Gaussian introduces a compact binary encoding and progressive training approach for 3D Gaussian segmentation, significantly reducing memory use and improving fine-grained segmentation accuracy.

Contribution

It proposes a novel binary encoding scheme and a progressive training strategy to enhance 3D Gaussian segmentation efficiency and detail.

Findings

Achieves state-of-the-art segmentation performance.

Reduces memory consumption substantially.

Speeds up inference process.

Abstract

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.