针对特定物体重建，大幅压缩模型体积并提升训练速度。
Object-Centric 2D Gaussian Splatting: Background Removal and Occlusion-Aware Pruning for Compact Object Models
- 用物体掩码实现精准目标重建，构建以对象为中心的模型。
- 通过感知遮挡的剪枝策略，模型体积减少96%，训练速度提升71%。
- 生成的模型可直接用于编辑与物理模拟，无需额外处理。
现有高斯点阵方法虽能有效重建完整场景，但无法聚焦特定物体，导致计算开销大，不适用于物体级应用。本文提出一种新方法，利用物体掩码实现目标导向重建，生成以对象为中心的模型。同时引入感知遮挡的剪枝策略，在不损失质量的前提下最小化高斯数量。所提方法重建出紧凑的物体模型，其高斯与网格表示相比基线最多缩小96%，训练速度最快提升71%，且保持良好质量。这些表示可直接用于外观编辑、物理仿真等下游任务，无需额外处理。
原文摘要 · Abstract (English)
Current Gaussian Splatting approaches are effective for reconstructing entire scenes but lack the option to target specific objects, making them computationally expensive and unsuitable for object-specific applications. We propose a novel approach that leverages object masks to enable targeted reconstruction, resulting in object-centric models. Additionally, we introduce an occlusion-aware pruning strategy to minimize the number of Gaussians without compromising quality. Our method reconstructs compact object models, yielding object-centric Gaussian and mesh representations that are up to 96% smaller and up to 71% faster to train compared to the baseline while retaining competitive quality. These representations are immediately usable for downstream applications such as appearance editing and physics simulation without additional processing.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。