通过空间自适应梯度重参数化，提升大卷积核在3D体积分析中的优化稳定性。
Spatially-Adaptive Gradient Re-parameterization for 3D Large Kernel Optimization
- 设计轻量调制网络生成感知感受野的缩放掩码，动态调整卷积核更新。
- 在五个3D分割数据集上均超越主流Transformer和固定先验方法。
- 无需复杂多分支结构，兼顾局部与全局收敛性，适合高分辨率3D任务。
大卷积核为高分辨率3D体数据分析提供了可扩展的替代方案，但直接增大核尺寸常导致优化不稳。基于有效感受野（ERFs）固有的空间偏置，我们理论上证明了结构化重参数化模块会引入空间变化的学习率，这对收敛至关重要。据此提出Rep3D框架，采用轻量调制网络生成感受野偏置的缩放掩码，在普通编码器架构中自适应重加权卷积核更新。该方法融合空间归纳偏置与优化感知学习，避免多分支设计复杂性，同时确保稳健的局部到全局收敛。在五个3D分割基准上的大量实验表明，Rep3D始终优于当前最先进的Transformer及固定先验基线。源代码已公开于https://github.com/leeh43/Rep3D。
原文摘要 · Abstract (English)
Large kernel convolutions offer a scalable alternative to vision transformers for high-resolution 3D volumetric analysis, yet naively increasing kernel size often leads to optimization instability. Motivated by the spatial bias inherent in effective receptive fields (ERFs), we theoretically demonstrate that structurally re-parameterized blocks induce spatially varying learning rates that are crucial for convergence. Leveraging this insight, we introduce Rep3D, a framework that employs a lightweight modulation network to generate receptive-biased scaling masks, adaptively re-weighting kernel updates within a plain encoder architecture. This approach unifies spatial inductive bias with optimization-aware learning, avoiding the complexity of multi-branch designs while ensuring robust local-to-global convergence. Extensive evaluations on five 3D segmentation benchmarks demonstrate that Rep3D consistently outperforms state-of-the-art transformer and fixed-prior baselines. The source code is publicly available at https://github.com/leeh43/Rep3D.
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。