PointTPA：面向3D场景理解的动态网络参数自适应
PointTPA: Dynamic Network Parameter Adaptation for 3D Scene Understanding
提出一种测试时参数自适应框架，通过序列化邻域分组和动态参数投影器为点云场景生成输入感知的骨干网络权重，可复用于其他3D场景理解任务的轻量级自适应。
原文依据 We propose PointTPA, a Test-time Parameter Adaptation framework that generates input-aware network parameters for scene-level point clouds.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论场景级点云理解由于几何形状多样、类别分布不均衡以及空间布局高度多变而仍具挑战性。现有方法提升了对象级性能，但在推理时依赖静态网络参数，限制了其对动态场景数据的适应性。我们提出PointTPA，一种测试时参数自适应框架，为场景级点云生成输入感知的网络参数。PointTPA采用基于序列化的邻域分组（SNG）形成局部连贯的块，并采用动态参数投影器（DPP）生成逐块自适应权重，使骨干网络能够根据场景特定变化调整其行为，同时保持较低的参数开销。集成到PTv3结构中，PointTPA通过引入两个轻量级模块（小于骨干网络参数的2%）展示了强大的参数效率。尽管参数开销极小，PointTPA在ScanNet验证集上达到78.4%的平均交并比，在多个基准上超越现有的参数高效微调（PEFT）方法，突显了我们的测试时动态网络参数自适应机制在增强3D场景理解方面的有效性。代码见https://github.com/H-EmbodVis/PointTPA。
英文原文摘要 1309 字符
Scene-level point cloud understanding remains challenging due to diverse geometries, imbalanced category distributions, and highly varied spatial layouts. Existing methods improve object-level performance but rely on static network parameters during inference, limiting their adaptability to dynamic scene data. We propose PointTPA, a Test-time Parameter Adaptation framework that generates input-aware network parameters for scene-level point clouds. PointTPA adopts a Serialization-based Neighborhood Grouping (SNG) to form locally coherent patches and a Dynamic Parameter Projector (DPP) to produce patch-wise adaptive weights, enabling the backbone to adjust its behavior according to scene-specific variations while maintaining a low parameter overhead. Integrated into the PTv3 structure, PointTPA demonstrates strong parameter efficiency by introducing two lightweight modules of less than 2% of the backbone's parameters. Despite this minimal parameter overhead, PointTPA achieves 78.4% mIoU on ScanNet validation, surpassing existing parameter-efficient fine-tuning (PEFT) methods across multiple benchmarks, highlighting the efficacy of our test-time dynamic network parameter adaptation mechanism in enhancing 3D scene understanding. The code is available at https://github.com/H-EmbodVis/PointTPA.