VIMCAN：基于混合 Mamba-交叉注意力网络的视觉-惯性三维人体姿态估计
VIMCAN: Visual-Inertial 3D Human Pose Estimation with Hybrid Mamba-Cross-Attention Network
用 Mamba 的选择性状态空间建模替代 Transformer 做时序建模、用交叉注意力补空间依赖，为视觉-惯性等多模态序列融合提供了一条线性复杂度、可实时（>60 FPS）的骨干设计思路，可直接迁移到其他需要长序列多模态融合的视觉任务。
原文依据 combines the efficient sequence modeling of Mamba with the spatial reasoning of Cross-Attention
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论深度学习的快速发展显著提升了多模态三维人体姿态估计（HPE）的精度。然而，最优（SOTA）的人体姿态估计流程仍依赖 Transformer，其二次复杂度使长序列的实时处理难以实现。Mamba 通过选择性状态空间建模解决了这一问题，在不牺牲表征能力的前提下实现了高效的序列处理。但它在多模态设定下难以捕捉复杂的空间依赖关系。为弥补这一差距，我们提出 VIMCAN，一种混合架构，将 Mamba 的高效序列建模与交叉注意力的空间推理相结合，在 RGB 关键点与可穿戴 IMU 数据之间进行鲁棒的视觉-惯性融合与人体姿态估计。通过利用 Mamba 的动态参数化进行时序建模、利用注意力提取空间依赖，VIMCAN 取得了更优的精度，在 TotalCapture 上的平均每关节位置误差（MPJPE）为 17.2 mm，在 3DPW 上为 45.3 mm。VIMCAN 优于此前基于 Transformer 的方法及其他最优方法，同时在消费级硬件上支持超过 60 帧/秒的实时推理。源代码将在 GitHub 上发布。
英文原文摘要 1258 字符
The rapid advances in deep learning have significantly enhanced the accuracy of multimodal 3D human pose estimation (HPE). However, the state-of-the-art (SOTA) HPE pipelines still rely on Transformers, whose quadratic complexity makes real-time processing for long sequences impractical. Mamba addresses this issue through selective state-space modeling, enabling efficient sequence processing without sacrificing representational power. Nevertheless, it struggles to capture complex spatial dependencies in multimodal settings. To bridge this gap, we propose VIMCAN, a hybrid architecture that combines the efficient sequence modeling of Mamba with the spatial reasoning of Cross-Attention, and performs robust visual-inertial fusion and human pose estimation between RGB keypoints and wearable IMU data. By leveraging Mamba's dynamic parameterization for temporal modeling and Attention for spatial dependency extraction, VIMCAN achieves superior accuracy, with mean per-joint position errors (MPJPE) of 17.2 mm on TotalCapture and 45.3 mm on 3DPW. VIMCAN outperforms prior Transformer-based and other SOTA approaches while supporting real-time inference at over 60 frames per second on consumer-grade hardware. The source code will be available on GitHub.