对于基于3D骨架的运动预测的并行多阶段整顿网络
Jianqi Zhong1,2, Conghui Ye1,2, Wenming Cao1,2
1Guangdong Key Laboratory of Intelligent Information Processing, Shenzhen, 518060, China.
Scientific reports
|October 30, 2024
概括
这项研究引入了一种新的三阶段并行网络,通过克服反复神经网络 (RNN) 错误积累来改善人类运动预测. 新模型增强了时空序列建模,以提供更准确的短期和长期预测.
科学领域:
- 计算机视觉 计算机视觉
- 机器学习 机器学习
- 人工智能的人工智能
背景情况:
- 循环神经网络 (RNN) 在人类预测任务的时空序列建模方面表现出色.
- RNN的逐步预测机制导致错误的积累,限制了运动预测的准确性.
- 现有的方法在准确的长期人类运动预测方面扎.
研究的目的:
- 为人类运动预测开发一个改进的网络架构.
- 为了减轻传统基于RNN的方法固有的错误积累问题.
- 为了提高短期和长期人类运动预测的准确性.
主要方法:
- 提出了三阶段并行预测网络,以指导具有明确目标的输出生成.
- 在并行网络中利用高维信息融合来生成最终输出.
- 设计了一个融合块,其中包含Gated Recurrent Unit (GRU) 和注意力机制,用于高效的高维特征提取.
主要成果:
- 拟议的三级并行网络显著超过了人类运动预测的最新方法.
- 在包括人类3.6M,CMU Mocap和3DPW在内的基准数据集上实现了卓越的性能.
- 在短期和长期运动预测场景中表现出有效性.
结论:
- 新的三级并行网络有效地解决了运动预测中的RNN错误积累.
- 拟议的架构增强了时空建模,以便更准确地预测人类运动.
- 这种方法代表了人类运动预测准确性和稳健性的重大进步.
相关概念视频
01:25Relative Motion Analysis using Rotating Axes
Consider a component AB undergoing a linear motion. Along with a linear motion, point B also rotates around point A. To comprehend this complex movement, position vectors for both points A and B are established using a stationary reference frame.
However, to express the relative position of point B relative to point A, an additional frame of reference, denoted as x'y', is necessary. This additional frame not only translates but also rotates relative to the fixed frame, making it instrumental in...
However, to express the relative position of point B relative to point A, an additional frame of reference, denoted as x'y', is necessary. This additional frame not only translates but also rotates relative to the fixed frame, making it instrumental in...
01:23Curvilinear Motion: Rectangular Components
Curvilinear motion characterizes the movement of a particle or object along a curved path, notably evident when envisioning a car navigating a winding road. If the car starts at point A, its position vector is established within a fixed frame of reference, where the ratio of the position vector to its magnitude signifies the unit vector pointing in the position vector's direction.
As the car advances, its position evolves over time. Quantifying the car's velocity involves computing the time...
As the car advances, its position evolves over time. Quantifying the car's velocity involves computing the time...