用专家混合模型提升3D重建边界清晰度，效果好且计算轻量。
MoE3D: A Mixture-of-Experts Module for 3D Reconstruction
- 通过多专家融合预测，动态选择每像素最优深度估计。
- 在深度不连续处显著减少模糊伪影，整体精度提升明显。
- 仅增加微量推理开销，小数据集上也能有效泛化。
我们提出一种简单而有效的前馈3D重建性能增强方法。现有方法在深度不连续区域表现不佳，因标准回归损失倾向于空间平均，导致边界模糊。为此，我们引入专家混合架构，通过组合多个平滑深度预测来处理深度边界的不确定性。一个Softmax加权头在像素级动态选择各假设。将该混合模型集成到预训练的先进3D模型中，显著减少了边界伪影并提升了整体重建精度。值得注意的是，该方法计算效率高，在少量训练数据上微调后仍能实现通用改进，且推理开销几乎无增加，为轻量化、高精度3D重建提供了新方向。
原文摘要 · Abstract (English)
We propose a simple yet effective approach to enhance the performance of feed-forward 3D reconstruction models. Existing methods often struggle near depth discontinuities, where standard regression losses encourage spatial averaging and thus blur sharp boundaries. To address this issue, we introduce a mixture-of-experts formulation that handles uncertainty at depth boundaries by combining multiple smooth depth predictions. A softmax weighting head dynamically selects among these hypotheses on a per-pixel basis. By integrating our mixture model into a pre-trained state-of-the-art 3D model, we achieve a substantial reduction of boundary artifacts and gains in overall reconstruction accuracy. Notably, our approach is highly compute efficient, delivering generalizable improvements even when fine-tuned on a small subset of training data while incurring only negligible additional inference computation, suggesting a promising direction for lightweight and accurate 3D reconstruction.
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。