通过多模态贝叶斯分布学习实现点云分析自适应
Adapting Point Cloud Analysis via Multimodal Bayesian Distribution Learning
把每个类别的文本先验与流式视觉特征都建模为高斯分布，用贝叶斯模型平均按后验证据自动加权融合，替代 cache 式 TTA 的启发式 logit 融合，可直接迁移到点云分类/分割等任务在分布偏移下的在线自适应。
原文依据 The two modalities are fused via Bayesian model averaging, which automatically adjusts their contributions based on posterior evidence.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论大型多模态 3D 视觉语言模型在多种 3D 任务上展现出强泛化能力，但在域偏移下性能仍会下降。这推动了近期关于测试时自适应（TTA）的研究，使模型能够利用测试时数据在线自适应。在现有 TTA 方法中，基于缓存的机制被广泛采用，以利用先前观察到的样本进行在线预测修正。然而，它们存储的历史信息有限，随着测试流演进会导致渐进式信息损失。此外，其预测 logits 以启发式方式融合，使自适应不稳定。为解决这些局限，我们提出 BayesMM，一个用于测试时点云分析的多模态贝叶斯分布学习框架。BayesMM 将每个类别的文本先验与流式视觉特征建模为高斯分布。文本参数由语义提示导出，而视觉参数随样本到达在线更新。两种模态通过贝叶斯模型平均进行融合，该方法依据后验证据自动调整二者的贡献。由此得到统一的预测，无需训练即可适应不断演化的测试时数据。在多个点云基准上的大量实验表明，BayesMM 在分布偏移下保持鲁棒性，平均提升超过 4%。
英文原文摘要 1394 字符
Large multimodal 3D vision-language models show strong generalization across diverse 3D tasks, but their performance still degrades under domain shifts. This has motivated recent studies on test-time adaptation (TTA), which enables models to adapt online using test-time data. Among existing TTA methods, cache-based mechanisms are widely adopted to leverage previously observed samples for online prediction refinement. However, they store limited historical information, leading to progressive information loss as the test stream evolves. In addition, their prediction logits are fused heuristically, making adaptation unstable. To address these limitations, we propose BayesMM, a multimodal Bayesian distribution learning framework for test-time point cloud analysis. BayesMM models textual priors and streaming visual features of each class as Gaussian distributions. Textual parameters are derived from semantic prompts, while visual parameters are updated online with arriving samples. The two modalities are fused via Bayesian model averaging, which automatically adjusts their contributions based on posterior evidence. This yields a unified prediction that adapts to evolving test-time data without training. Extensive experiments on multiple point cloud benchmarks demonstrate that BayesMM maintains robustness under distributional shifts, yielding over 4 percent average improvement.