实验证明,MotionAGForme在Human3.6M数据集和MPI-INF-3DHP数据集上使用的参数是之前领先模型的四分之一,计算效率是其三倍。 创新点: 引入了Attention-GCNFormer (AGFormer)模块,通过使用两个并行的Transformer和GCNFormer流,将通道数量分割。GCNFormer模块利用相邻关节之间的局部关系,输出与Transformer输出互补的新表示。
实验证明,MotionAGForme在Human3.6M数据集和MPI-INF-3DHP数据集上使用的参数是之前领先模型的四分之一,计算效率是其三倍。 创新点: 引入了Attention-GCNFormer (AGFormer)模块,通过使用两个并行的Transformer和GCNFormer流,将通道数量分割。GCNFormer模块利用相邻关节之间的局部关系,输出与Transformer输出互补的新表示。
尽管如此,涉及 GCNFormer 和 Transformer 的混合方法产生了显着的改进,与单独使用 Transformer 相比,P1 误差减少了 5.2mm。 此外,这两个模块的顺序融合不如它们的并行集成有效。 表7. 不同 MetaFormer 集成的比较。 所有模型均使用我们的 MotionAGFormer-B 设置在 Human3.6M 上进行训练。 4.6. Qualitative Analysi...
In the field of skeleton-based three-dimensional human motion prediction (HMP), recent advances in methods that model human motion through a series of intermediate states have demonstrated considerable potential. Nevertheless, a significant challenge with these approaches is the accurate acquisition of mu...
This is the official PyTorch implementation of the paper "MotionAGFormer: Enhancing 3D Human Pose Estimation With a Transformer-GCNFormer Network" (WACV 2024).EnvironmentThe project is developed under the following environment:Python 3.8.10 PyTorch 2.0.0 CUDA 12.2For installation of the project ...
With the development of information technology, massive traffic data-driven short-term traffic situation analysis of urban road networks has become a research hotspot in urban traffic management. Accurate vehicle trajectory and traffic flow prediction ca
Cross-scale cascade transformer for multimodal human action recognition 2023, Pattern Recognition Letters Citation Excerpt : Visual sensors like Kinect have been widely used and provide a variety of modalities [1,3], e.g., RGB, depth, infrared radiation (IR), skeletons, etc. which can compensate...
Decoupled spatio-temporal grouping transformer for skeleton-based action recognition 2024, Visual Computer Non-local Graph Convolutional Network 2024, Circuits, Systems, and Signal Processing Motion-Driven Spatial and Temporal Adaptive High-Resolution Graph Convolutional Networks for Skeleton-Based Action Recogn...
3.1 Pipeline Overview基于骨架的数据可以从运动捕获(Motion-Capture)设备中获取，也可以从视频中通过姿势...
Skeleton-based Action Recognition via Spatial and Temporal Transformer Networks 1 code implementation • 17 Aug 2020 Skeleton-based Human Activity Recognition has achieved great interest in recent years as skeleton data has demonstrated being robust to illumination changes, body scales, dynamic camera ...