解决类别间姿态估计优化冲突，提升多类别6D姿态预测精度
DecomPose: Disentangling Cross-Category Optimization Contention for Category-Level 6D Object Pose Estimation
- 按类别难易度动态分组，分路处理不同类别对应关系
- 简单类别用高容量分支稳定训练，复杂类别用轻量分支抑制噪声
- 在三个基准数据集上显著降低优化冲突，性能全面领先
类别级6D物体姿态估计通常被建模为多类别联合学习问题，采用完全共享的模型参数。然而，类别间显著的几何异质性导致共享模块中优化信号相互干扰，引发梯度冲突和负迁移。为此，我们首先引入基于梯度的诊断方法，量化模块级跨类别优化竞争程度。基于诊断结果，提出DecomPose框架：（1）基于难度感知的梯度解耦，利用数据驱动的难易度代理对类别分组，并将每类实例路由至特定对应分支，隔离不兼容更新；（2）基于稳定性的非对称分支设计，为结构简单的类别分配高容量分支作为稳定优化锚点，而对复杂类别使用轻量分支以抑制噪声更新，缓解负迁移。在REAL275、CAMERA25和HouseCat6D上的大量实验表明，DecomPose有效降低了跨类别优化竞争，多个基准上均取得更优的姿态估计性能。
原文摘要 · Abstract (English)
Category-level 6D object pose estimation is typically formulated as a multi-category joint learning problem with fully shared model parameters. However, pronounced geometric heterogeneity across categories entangles incompatible optimization signals in shared modules, resulting in gradient conflicts and negative transfer during training. To address this challenge, we first introduce gradient-based diagnostics to quantify module-level cross-category contention. Building on results of diagnostics, we propose DecomPose, a difficulty-aware decomposition framework that mitigates optimization contention via: (1) difficulty-aware gradient decoupling, which groups categories using a data-driven difficulty proxy and routes each instance to a group-specific correspondence branch to isolate incompatible updates; and (2) stability-driven asymmetric branching, which assigns higher-capacity branches to structurally simple categories as stable optimization anchors while constraining complex categories with lightweight branches to suppress noisy updates and alleviate negative transfer. Extensive experiments on REAL275, CAMERA25, and HouseCat6D demonstrate that DecomPose effectively reduces cross-category optimization contention and delivers superior pose estimation performance across multiple benchmarks.
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。